From bedouglas at earthlink.net  Thu Nov  1 00:45:07 2007
From: bedouglas at earthlink.net (bruce)
Date: Wed, 31 Oct 2007 20:45:07 -0800
Subject: [SciPy-user] SciPY/NumPY - rhel4 installation issues....
Message-ID: <067e01c81c41$fc024bc0$0301a8c0@tmesa.com>

Hi.

I'm trying to install numpy/scipy on a rhel system with python v2.4.3

I'm following the information from various sites. My goal is to be able to
successfully get numpy/scipy up/running, with a few basic tests.

I'm using the tar files for:
	drwxr-xr-x  6 root  root     4096 Oct 31 19:30 .
	drwxr-xr-x 32 root  root     4096 Oct 31 12:26 ..
	drwxr-xr-x 12 root  root     4096 Oct 31 18:29 ATLAS
	-rw-r--r--  1 root  root  3192987 Oct 31 12:35 atlas3.8.0.tar.gz
	drwxr-x---  8 13686  731     4096 Oct 31 18:11 lapack-3.1.1
	-rw-r--r--  1 root  root 10407595 Feb 26  2007 lapack-3.1.1.tgz
	drwxrwxr-x  4 test  test     4096 Oct 31 19:43 numpy-1.0.3.1
	-rw-r--r--  1 root  root  1500255 Oct 31 12:49 numpy-1.0.3.1.tar.gz
	-rw-r--r--  1 root  root       54 Oct 31 12:27 readme.txt
	drwxrwxr-x  3 test  test     4096 Sep 22 01:14 scipy-0.6.0
	-rw-r--r--  1 root  root  6572954 Oct 31 12:52 scipy-0.6.0.tar.gz

I believe I've successfully downloaded/built/installed the atlas/lapack
apps.

I'm using a RHEL4 system with a single 1.8GHz. processor.

I did a:
  "python setup.py install" for numpy, and it appeared to finish.

I then did a:
  "python setup.py install" for scipy, and got the following err:
	[root at mfgtest3 <Branch GodzillaTest1.1.2> scipy-0.6.0]# python setup.py
install
	Traceback (most recent call last):
	  File "setup.py", line 53, in ?
	    setup_package()
	  File "setup.py", line 28, in setup_package
	    from numpy.distutils.core import setup
	  File "/usr/lib/python2.4/site-packages/numpy/__init__.py", line 39, in ?
	    import core
	  File "/usr/lib/python2.4/site-packages/numpy/core/__init__.py", line 5,
in ?
	    import multiarray
	ImportError: /usr/lib/python2.4/site-packages/numpy/core/multiarray.so:
undefined symbol: 	PyUnicodeUCS2_FromUnicode

I can't find  anything via google to indicate what's the cause...

Any thoughts/comments would be seriously helpful!!

Could this be an issue of incompatible versions between numpy/scipy??

thanks

-sam




From parvel.gu at gmail.com  Thu Nov  1 02:08:38 2007
From: parvel.gu at gmail.com (Parvel Gu)
Date: Thu, 1 Nov 2007 14:08:38 +0800
Subject: [SciPy-user] Generating random variables in a joint normal
	distribution?
In-Reply-To: <4725637C.7030207@gmail.com>
References: <d195a2b10710281936o2d32e8d1w877f07d358d7fef5@mail.gmail.com>
	<4725487E.40509@gmail.com>
	<d195a2b10710282126h59104899l91d266e8b6b00a51@mail.gmail.com>
	<4725637C.7030207@gmail.com>
Message-ID: <d195a2b10710312308p66b4d01ahe22eb0d201ec9d2e@mail.gmail.com>

Hi,

Thanks so much for the reply. I could go furthur in this problem.

And here is the following issue:

In my understanding, the continues calls to random.multivariate_normal
would get a serials of random vairables which are all following the
joint normal distribution. Then what about how to reset the random
generating in each iteration?

Assuming that I have to simulate something for 50 iterations then to
get the average value. In each iteration I need a serial of random
pairs which follows the joint normal distrubition. The interations are
expected to be independent to each other.

Thus:

In iteration 1:
       serial = []
       in sub loop
          serial.append(random.multivariate_normal(m, cov))
       ...

And in iteration 2, I want a fresh serial which should not be affected
by the previous one        s. Is there any problem to empty the serial
then still call  random.multivariate_normal(m, cov) ?
In my understanding if there is no refresh or something performed to
the internal random generator, the random.multivariate_normal would
continue to generate variables in the context of the previous ones.

Br,
Parvel


From robert.kern at gmail.com  Thu Nov  1 04:09:00 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 01 Nov 2007 03:09:00 -0500
Subject: [SciPy-user] Generating random variables in a joint
	normal	distribution?
In-Reply-To: <d195a2b10710312308p66b4d01ahe22eb0d201ec9d2e@mail.gmail.com>
References: <d195a2b10710281936o2d32e8d1w877f07d358d7fef5@mail.gmail.com>	<4725487E.40509@gmail.com>	<d195a2b10710282126h59104899l91d266e8b6b00a51@mail.gmail.com>	<4725637C.7030207@gmail.com>
	<d195a2b10710312308p66b4d01ahe22eb0d201ec9d2e@mail.gmail.com>
Message-ID: <4729899C.8090904@gmail.com>

Parvel Gu wrote:
> Hi,
> 
> Thanks so much for the reply. I could go furthur in this problem.
> 
> And here is the following issue:
> 
> In my understanding, the continues calls to random.multivariate_normal
> would get a serials of random vairables which are all following the
> joint normal distribution. Then what about how to reset the random
> generating in each iteration?
> 
> Assuming that I have to simulate something for 50 iterations then to
> get the average value. In each iteration I need a serial of random
> pairs which follows the joint normal distrubition. The interations are
> expected to be independent to each other.
> 
> Thus:
> 
> In iteration 1:
>        serial = []
>        in sub loop
>           serial.append(random.multivariate_normal(m, cov))
>        ...
> 
> And in iteration 2, I want a fresh serial which should not be affected
> by the previous one        s. Is there any problem to empty the serial
> then still call  random.multivariate_normal(m, cov) ?
> In my understanding if there is no refresh or something performed to
> the internal random generator, the random.multivariate_normal would
> continue to generate variables in the context of the previous ones.

Just keep sampling. The samples will be as independent of each other as you can
possibly get them. Reseeding the generator (the only thing I can imagine that
you meant by "refresh or something") will only make things worse.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From lxander.m at gmail.com  Thu Nov  1 09:13:28 2007
From: lxander.m at gmail.com (Alexander Michael)
Date: Thu, 1 Nov 2007 09:13:28 -0400
Subject: [SciPy-user] SciPY/NumPY - rhel4 installation issues....
In-Reply-To: <067e01c81c41$fc024bc0$0301a8c0@tmesa.com>
References: <067e01c81c41$fc024bc0$0301a8c0@tmesa.com>
Message-ID: <525f23e80711010613g73959bdaw98d31f1288f75654@mail.gmail.com>

On 11/1/07, bruce <bedouglas at earthlink.net> wrote:
> I then did a:
>   "python setup.py install" for scipy, and got the following err:
>         [root at mfgtest3 <Branch GodzillaTest1.1.2> scipy-0.6.0]# python setup.py
> install
>         Traceback (most recent call last):
>           File "setup.py", line 53, in ?
>             setup_package()
>           File "setup.py", line 28, in setup_package
>             from numpy.distutils.core import setup
>           File "/usr/lib/python2.4/site-packages/numpy/__init__.py", line 39, in ?
>             import core
>           File "/usr/lib/python2.4/site-packages/numpy/core/__init__.py", line 5,
> in ?
>             import multiarray
>         ImportError: /usr/lib/python2.4/site-packages/numpy/core/multiarray.so:
> undefined symbol:       PyUnicodeUCS2_FromUnicode
>
> I can't find  anything via google to indicate what's the cause...
>
> Any thoughts/comments would be seriously helpful!!

This could be a result of your numpy installation being incomplete
(likely do to not being correctly linked to blas/lapack). What happens
when you try:

python -c "import numpy"

on its own? Does it fail with the same error? If so, you need to focus
on numpy. One subtlety to be aware of is that you need to remove the
build directory in your numpy directory whenever you need to rebuild
numpy.

HTH,
Alex


From berthe.loic at gmail.com  Thu Nov  1 10:27:38 2007
From: berthe.loic at gmail.com (LB)
Date: Thu, 01 Nov 2007 07:27:38 -0700
Subject: [SciPy-user] Bug in scipy.integrate.odeint ?
Message-ID: <1193927258.315014.184820@50g2000hsm.googlegroups.com>

Hi,

I find the behaviour of scipy.integrate.odeint very strange :
>>> from numpy import *
>>> from scipy import integrate
>>>
>>> def f(x, t):
...        return cos(x)
...
>>>
>>> X0 = array([0.0])
>>> X0_ini = X0.copy()
>>> t = linspace(0, 20., 200)
>>> X = integrate.odeint(f, X0, t)
>>> X0 == X0_ini
array([False], dtype=bool)
>>> X0 == X[-1]
array([True], dtype=bool)

Why do odeint modify X0 ?
Is there any reason for this ?
For me, it seems really like a bug and could easily lead to bugs if
there's no mention of this in the doc string.

--
LB



From cookedm at physics.mcmaster.ca  Thu Nov  1 10:56:19 2007
From: cookedm at physics.mcmaster.ca (David M. Cooke)
Date: Thu, 1 Nov 2007 10:56:19 -0400
Subject: [SciPy-user] Bug in scipy.integrate.odeint ?
In-Reply-To: <1193927258.315014.184820@50g2000hsm.googlegroups.com>
References: <1193927258.315014.184820@50g2000hsm.googlegroups.com>
Message-ID: <CFDBDC2D-5F0F-47A1-BA1A-7965127BC671@physics.mcmaster.ca>

On Nov 1, 2007, at 10:27 , LB wrote:

> Hi,
>
> I find the behaviour of scipy.integrate.odeint very strange :
>>>> from numpy import *
>>>> from scipy import integrate
>>>>
>>>> def f(x, t):
> ...        return cos(x)
> ...
>>>>
>>>> X0 = array([0.0])
>>>> X0_ini = X0.copy()
>>>> t = linspace(0, 20., 200)
>>>> X = integrate.odeint(f, X0, t)
>>>> X0 == X0_ini
> array([False], dtype=bool)
>>>> X0 == X[-1]
> array([True], dtype=bool)
>
> Why do odeint modify X0 ?
> Is there any reason for this ?
> For me, it seems really like a bug and could easily lead to bugs if
> there's no mention of this in the doc string.


I'm guessing you've got an old version of scipy installed, as it was  
changed in February to use a copy of the input arguments instead.

-- 
|>|\/|<
/------------------------------------------------------------------\
|David M. Cooke              http://arbutus.physics.mcmaster.ca/dmc/
|cookedm at physics.mcmaster.ca



From bradford.n.cross at gmail.com  Thu Nov  1 13:29:16 2007
From: bradford.n.cross at gmail.com (Bradford Cross)
Date: Thu, 1 Nov 2007 10:29:16 -0700
Subject: [SciPy-user] Installing TimeSeries
Message-ID: <ea7d6a710711011029s2182dc13j16f2377e95cb2be1@mail.gmail.com>

Install works OK for MAskedArray but fails for TimeSeries with following
output:

bcross at ubuntu:~/maskedarray$ sudo python setup.py install
Password:
running install
running build
running config_fc
running build_py
running install_lib
running install_data
running install_egg_info
Removing /usr/lib/python2.5/site-packages/maskedarray-0.0.0.egg-info
Writing /usr/lib/python2.5/site-packages/maskedarray-0.0.0.egg-info
bcross at ubuntu:~/maskedarray$ cd
bcross at ubuntu:~$ cd timeseries/
bcross at ubuntu:~/timeseries$ sudo python setup.py install
Warning: Assuming default configuration (./lib/{setup_lib,setup}.py was not
found)
Appending timeseries.lib configuration to timeseries
Ignoring attempt to set 'name' (from 'timeseries' to 'timeseries.lib')
Warning: Assuming default configuration (./tests/{setup_tests,setup}.py was
not found)
Appending timeseries.tests configuration to timeseries
Ignoring attempt to set 'name' (from 'timeseries' to 'timeseries.tests')
running install
running build
running config_fc
running build_src
building extension "timeseries.cseries" sources
running build_py
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
building 'timeseries.cseries' extension
compiling C sources
C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O2 -Wall
-Wstrict-prototypes -fPIC

compile options:
'-I/usr/lib/python2.5/site-packages/numpy/core/include/numpy -I./include
-I/usr/lib/python2.5/site-packages/numpy/core/include
-I/usr/include/python2.5 -c'
gcc: ./src/cseries.c
In file included from ./src/cseries.c:1:
./include/c_lib.h:4:20: error: Python.h: No such file or directory
./include/c_lib.h:5:26: error: structmember.h: No such file or directory
In file included from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/arrayobject.h:14,
                 from ./include/c_lib.h:6,
                 from ./src/cseries.c:1:
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:91:2:
error: #error Must use Python with unicode enabled.
In file included from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/arrayobject.h:14,
                 from ./include/c_lib.h:6,
                 from ./src/cseries.c:1:
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:918:
error: expected '=', ',', ';', 'asm' or '__attribute__' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:919:
error: expected '=', ',', ';', 'asm' or '__attribute__' before 'npy_uintp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1009:
error: expected '=', ',', ';', 'asm' or '__attribute__' before '*' token
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1010:
error: expected ')' before '*' token
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1012:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1012:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1013:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1023:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1023:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1025:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1025:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1026:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1028:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1031:
error: expected ')' before '*' token
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1034:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1036:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1037:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1037:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1039:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1044:
error: expected specifier-qualifier-list before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1057:
error: expected specifier-qualifier-list before 'PyArray_GetItemFunc'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1159:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1191:
error: expected specifier-qualifier-list before 'PyObject'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1201:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1227:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1234:
error: expected declaration specifiers or '...' before 'PyObject'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1367:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1505:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1562:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1762:
error: expected specifier-qualifier-list before 'npy_intp'
In file included from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1777,
                 from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/arrayobject.h:14,
                 from ./include/c_lib.h:6,
                 from ./src/cseries.c:1:
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:397:
error: 'NULL' undeclared here (not in a function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:
In function '_import_array':
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: 'PyObject' undeclared (first use in this function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: (Each undeclared identifier is reported only once
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: for each function it appears in.)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: 'numpy' undeclared (first use in this function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
warning: implicit declaration of function 'PyImport_ImportModule'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:946:
error: 'c_api' undeclared (first use in this function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:948:
warning: implicit declaration of function 'PyObject_GetAttrString'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:949:
warning: implicit declaration of function 'Py_DECREF'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:950:
warning: implicit declaration of function 'PyCObject_Check'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:951:
warning: implicit declaration of function 'PyCObject_AsVoidPtr'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:958:
warning: implicit declaration of function 'PyErr_Format'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:958:
error: 'PyExc_RuntimeError' undeclared (first use in this function)
In file included from ./src/cseries.c:1:
./include/c_lib.h: At top level:
./include/c_lib.h:19: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:20: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:21: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:22: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:23: error: expected ')' before '*' token
./include/c_lib.h:24: error: expected ')' before '*' token
./include/c_lib.h:26: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:28: error: expected ')' before '*' token
In file included from ./src/cseries.c:2:
./include/c_dates.h:103: error: expected ')' before '*' token
./include/c_dates.h:109: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:110: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:113: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:114: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:115: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:116: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:118: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:119: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:121: error: expected ')' before '*' token
In file included from ./src/cseries.c:3:
./include/c_tseries.h:6: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:8: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:9: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:10: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:11: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:13: error: expected ')' before '*' token
./src/cseries.c:5: error: expected '=', ',', ';', 'asm' or '__attribute__'
before 'cseries_methods'
./src/cseries.c:55: error: expected '=', ',', ';', 'asm' or '__attribute__'
before 'initcseries'
In file included from ./src/cseries.c:1:
./include/c_lib.h:4:20: error: Python.h: No such file or directory
./include/c_lib.h:5:26: error: structmember.h: No such file or directory
In file included from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/arrayobject.h:14,
                 from ./include/c_lib.h:6,
                 from ./src/cseries.c:1:
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:91:2:
error: #error Must use Python with unicode enabled.
In file included from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/arrayobject.h:14,
                 from ./include/c_lib.h:6,
                 from ./src/cseries.c:1:
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:918:
error: expected '=', ',', ';', 'asm' or '__attribute__' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:919:
error: expected '=', ',', ';', 'asm' or '__attribute__' before 'npy_uintp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1009:
error: expected '=', ',', ';', 'asm' or '__attribute__' before '*' token
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1010:
error: expected ')' before '*' token
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1012:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1012:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1013:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1023:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1023:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1025:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1025:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1026:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1028:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1031:
error: expected ')' before '*' token
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1034:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1036:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1037:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1037:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1039:
error: expected declaration specifiers or '...' before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1044:
error: expected specifier-qualifier-list before 'npy_intp'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1057:
error: expected specifier-qualifier-list before 'PyArray_GetItemFunc'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1159:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1191:
error: expected specifier-qualifier-list before 'PyObject'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1201:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1227:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1234:
error: expected declaration specifiers or '...' before 'PyObject'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1367:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1505:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1562:
error: expected specifier-qualifier-list before 'PyObject_HEAD'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1762:
error: expected specifier-qualifier-list before 'npy_intp'
In file included from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/ndarrayobject.h:1777,
                 from
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/arrayobject.h:14,
                 from ./include/c_lib.h:6,
                 from ./src/cseries.c:1:
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:397:
error: 'NULL' undeclared here (not in a function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:
In function '_import_array':
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: 'PyObject' undeclared (first use in this function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: (Each undeclared identifier is reported only once
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: for each function it appears in.)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
error: 'numpy' undeclared (first use in this function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:945:
warning: implicit declaration of function 'PyImport_ImportModule'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:946:
error: 'c_api' undeclared (first use in this function)
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:948:
warning: implicit declaration of function 'PyObject_GetAttrString'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:949:
warning: implicit declaration of function 'Py_DECREF'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:950:
warning: implicit declaration of function 'PyCObject_Check'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:951:
warning: implicit declaration of function 'PyCObject_AsVoidPtr'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:958:
warning: implicit declaration of function 'PyErr_Format'
/usr/lib/python2.5/site-packages/numpy/core/include/numpy/__multiarray_api.h:958:
error: 'PyExc_RuntimeError' undeclared (first use in this function)
In file included from ./src/cseries.c:1:
./include/c_lib.h: At top level:
./include/c_lib.h:19: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:20: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:21: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:22: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:23: error: expected ')' before '*' token
./include/c_lib.h:24: error: expected ')' before '*' token
./include/c_lib.h:26: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_lib.h:28: error: expected ')' before '*' token
In file included from ./src/cseries.c:2:
./include/c_dates.h:103: error: expected ')' before '*' token
./include/c_dates.h:109: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:110: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:113: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:114: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:115: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:116: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:118: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:119: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_dates.h:121: error: expected ')' before '*' token
In file included from ./src/cseries.c:3:
./include/c_tseries.h:6: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:8: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:9: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:10: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:11: error: expected '=', ',', ';', 'asm' or
'__attribute__' before '*' token
./include/c_tseries.h:13: error: expected ')' before '*' token
./src/cseries.c:5: error: expected '=', ',', ';', 'asm' or '__attribute__'
before 'cseries_methods'
./src/cseries.c:55: error: expected '=', ',', ';', 'asm' or '__attribute__'
before 'initcseries'
error: Command "gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O2 -Wall
-Wstrict-prototypes -fPIC
-I/usr/lib/python2.5/site-packages/numpy/core/include/numpy -I./include
-I/usr/lib/python2.5/site-packages/numpy/core/include
-I/usr/include/python2.5 -c ./src/cseries.c -o
build/temp.linux-i686-2.5/src/cseries.o"
failed with exit status 1
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071101/ac74fefb/attachment.html>

From zelbier at gmail.com  Thu Nov  1 14:06:29 2007
From: zelbier at gmail.com (Olivier)
Date: Thu, 01 Nov 2007 18:06:29 -0000
Subject: [SciPy-user] Bug in scipy.sparse
In-Reply-To: <b2da89060710310401r5d8b407m70d10048e08937cb@mail.gmail.com>
References: <b2da89060710310401r5d8b407m70d10048e08937cb@mail.gmail.com>
Message-ID: <1193940389.563473.128880@d55g2000hsg.googlegroups.com>

Nobody seems to care much about sparse arrays in scipy.... One reason
why i tremendously prefer arrays over matrices is that it is extremely
easy to multiply all the rows or columns of a matrix by a vector of
coefficients. You just do M*V to multiply column-wise or
M*V.reshape(-1,1) to multiply row-wise.

I hate having to create a diagonal matrix for such a simple task.

Nobody is interested in having sparse arrays in scipy?


== Olivier

On Oct 31, 12:01 pm, "Olivier Verdier" <zelb... at gmail.com> wrote:
> The first thing I'm surprised about with scipy.sparse is that it uses the
> matrix type instead of the array type. This is very unfortunate, in my
> opinion.
> The bug is that a sparse matrix won't work correctly with `dot` and `array`:
>
> spmat = speye(3,3)
> vec  = array([1,0,0])
>
> dot(spmat, vec)
> returns:
> array([  (0, 0) 1.0
>   (1, 1)        1.0
>   (2, 2)        1.0,
>          (0, 0) 0.0
>   (1, 1)        0.0
>   (2, 2)        0.0,
>          (0, 0) 0.0
>   (1, 1)        0.0
>   (2, 2)        0.0], dtype=object)
>
> Ouch!!
>
> The right result may however be obtained with spmat * vec.
>
> This is a pity because any code that works with arrays or matrices in
> general and uses `dot` to do the matrix vector multiplication will be
> *broken* with sparse matrices.
>
> How reliable is scipy.sparse? Is there any plan to make it more compatible
> with the array type? Behave like the array type? How can I help?
>
> == Olivier
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-u... at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user



From s.mientki at ru.nl  Thu Nov  1 14:16:54 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Thu, 01 Nov 2007 19:16:54 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
Message-ID: <472A1816.6060902@ru.nl>



Matthieu Brucher wrote:
>
>
> 2007/10/31, Paul Barrett <pebarrett at gmail.com 
> <mailto:pebarrett at gmail.com>>:
>
>     No, you are not dreaming.  I also commented on this possibility
>     several years ago to some people in the SciPy community.
>     Unfortunately, I have not had the time or motivation to act on it.  It
>     would certainly be nice to have a basic package to show people.
>     You'll need to create a set of APIs so that you can connect to
>     instrumentation.  It shouldn't be too hard now that the interfaces
>     have been standardized.
>
>
> Sorry, it was in fact Vision and Viper I saw in the past as well.
Where can I find Viper ?
Google hates those common names ;-)
> The idea is great but the application, IMHO, is not quite optimal :
> - even for the tutorial, the rotate bloc should have two inputs, the 
> image and the rotation angle (this last factor could be given by the 
> result of another algorithm)
maybe 3, also one for the interpolation method ?
I would also make the major controls (for models with not too many 
controls),
in this case the rotation angle directly available with a small slider 
on the visual component.
> - what about multithreading in this model ?
> - what about flow control ? (which is really a huge issue that is 
> lacking in ITK for instance)
> - is there a way to create a bunch of blocks easily ?
>
Does LabView supports these ?
> All this has consequences on the interface. Besides, it would be very 
> great to use Traits as input and output blocks.
I Don't know Traits, so can't judge that.

cheers,
Stef



From millman at berkeley.edu  Thu Nov  1 14:19:52 2007
From: millman at berkeley.edu (Jarrod Millman)
Date: Thu, 1 Nov 2007 11:19:52 -0700
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472A1816.6060902@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<472A1816.6060902@ru.nl>
Message-ID: <c7009a550711011119x6f0651cavdc18d43371042698@mail.gmail.com>

On Nov 1, 2007 11:16 AM, Stef Mientki <s.mientki at ru.nl> wrote:
> Where can I find Viper ?

http://www.scripps.edu/~sanner/python/viper/index.html

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From matthieu.brucher at gmail.com  Thu Nov  1 14:56:47 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 1 Nov 2007 19:56:47 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472A1816.6060902@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<472A1816.6060902@ru.nl>
Message-ID: <e76aa17f0711011156m432d97ecq7fe80169ea37114d@mail.gmail.com>

>
> maybe 3, also one for the interpolation method ?
> I would also make the major controls (for models with not too many
> controls),
> in this case the rotation angle directly available with a small slider
> on the visual component.



Exactly.


> - what about multithreading in this model ?
> > - what about flow control ? (which is really a huge issue that is
> > lacking in ITK for instance)
> > - is there a way to create a bunch of blocks easily ?
>
> Does LabView supports these ?



I don't know, but I know that it is what I need and what we want to build
more or less in the lab I'm doing my PhD. The algorithms are very long to
run, and there is multiple calculations that can be run in parallel. This
induces multithreading, flow control (actually not very complicated once a
flow control instruction is thought as another block) and easiness of
creation (perhaps a simple xml file that would be parsed to create a
meta-class that would be then instanciated).
And something else is relevent : saving results after each block.

-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071101/20186470/attachment.html>

From stef.mientki at gmail.com  Thu Nov  1 15:04:20 2007
From: stef.mientki at gmail.com (stef mientki)
Date: Thu, 01 Nov 2007 20:04:20 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <c7009a550711011119x6f0651cavdc18d43371042698@mail.gmail.com>
References: <4728FFC8.2070007@ru.nl>	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>	<472A1816.6060902@ru.nl>
	<c7009a550711011119x6f0651cavdc18d43371042698@mail.gmail.com>
Message-ID: <472A2334.3080109@gmail.com>

Jarrod Millman wrote:
> On Nov 1, 2007 11:16 AM, Stef Mientki <s.mientki at ru.nl> wrote:
>   
>> Where can I find Viper ?
>>     
>
> http://www.scripps.edu/~sanner/python/viper/index.html
>
>   
if  Vision is Viper :
  choice = False
  :-)

cheers,
Stef



From millman at berkeley.edu  Thu Nov  1 15:13:04 2007
From: millman at berkeley.edu (Jarrod Millman)
Date: Thu, 1 Nov 2007 12:13:04 -0700
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472A2334.3080109@gmail.com>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<472A1816.6060902@ru.nl>
	<c7009a550711011119x6f0651cavdc18d43371042698@mail.gmail.com>
	<472A2334.3080109@gmail.com>
Message-ID: <c7009a550711011213l59b96f3cm869b7c96c62f8240@mail.gmail.com>

On Nov 1, 2007 12:04 PM, stef mientki <stef.mientki at gmail.com> wrote:
> Jarrod Millman wrote:
> > On Nov 1, 2007 11:16 AM, Stef Mientki <s.mientki at ru.nl> wrote:
> >> Where can I find Viper ?
> >>
> > http://www.scripps.edu/~sanner/python/viper/index.html
> >
> if  Vision is Viper :
>   choice = False
>   :-)

I believe Viper was renamed Vision.  You can see the legacy of Viper
still in the URL path and it is referenced in their publications at
the bottom of the page.

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From mattknox_ca at hotmail.com  Thu Nov  1 15:21:06 2007
From: mattknox_ca at hotmail.com (Matt Knox)
Date: Thu, 1 Nov 2007 19:21:06 +0000 (UTC)
Subject: [SciPy-user] Installing TimeSeries
References: <ea7d6a710711011029s2182dc13j16f2377e95cb2be1@mail.gmail.com>
Message-ID: <loom.20071101T191614-513@post.gmane.org>

> ./include/c_lib.h:4:20: error: Python.h: No such file or directory

I am mostly a windows guy, but I *believe* you may need to install some python
development package (which is probably available through the standard package
repository menu in Ubuntu). Might be called something like "python-devel".

If that doesn't work, perhaps one of the many linux folks on this mailing list
could help you.

Once you get it working, let me know if you need any help using it or have any
suggestions, etc.

- Matt




From stefan at sun.ac.za  Thu Nov  1 16:13:02 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Thu, 1 Nov 2007 22:13:02 +0200
Subject: [SciPy-user] Bug in scipy.sparse
In-Reply-To: <1193940389.563473.128880@d55g2000hsg.googlegroups.com>
References: <b2da89060710310401r5d8b407m70d10048e08937cb@mail.gmail.com>
	<1193940389.563473.128880@d55g2000hsg.googlegroups.com>
Message-ID: <20071101201302.GB14429@mentat.za.net>

On Thu, Nov 01, 2007 at 06:06:29PM -0000, Olivier wrote:
> Nobody seems to care much about sparse arrays in scipy.... One reason
> why i tremendously prefer arrays over matrices is that it is extremely
> easy to multiply all the rows or columns of a matrix by a vector of
> coefficients. You just do M*V to multiply column-wise or
> M*V.reshape(-1,1) to multiply row-wise.
> 
> I hate having to create a diagonal matrix for such a simple task.
> 
> Nobody is interested in having sparse arrays in scipy?

It's not that we aren't interested, but someone has to write it.  The
sparse matrix base classes need to support dot, element-wise
multiplication and broadcasting.  Then it is a simple matter of
subclassing and overriding * to get matrix functionality.  Nathan has
really done a lot to improve the underlying engine, but the Python end
of things could do with a good refactoring.

Cheers
St?fan


From nmarais at sun.ac.za  Fri Nov  2 09:51:13 2007
From: nmarais at sun.ac.za (Neilen Marais)
Date: Fri, 2 Nov 2007 13:51:13 +0000 (UTC)
Subject: [SciPy-user] Bug in scipy.sparse
References: <b2da89060710310401r5d8b407m70d10048e08937cb@mail.gmail.com>
Message-ID: <fgfa0h$uji$1@ger.gmane.org>

Hi Olivier

On Wed, 31 Oct 2007 12:01:50 +0100, Olivier Verdier wrote:

> The first thing I'm surprised about with scipy.sparse is that it uses
> the matrix type instead of the array type. This is very unfortunate, in
> my opinion.
> The bug is that a sparse matrix won't work correctly with `dot` and
> `array`:
> 
> This is a pity because any code that works with arrays or matrices in
> general and uses `dot` to do the matrix vector multiplication will be
> *broken* with sparse matrices.

Yeah, I agree this is a bit irritating sometimes. OTOH you usually know 
when you are dealing with sparse matrices since they tend to be fairly 
specific in their application.

> 
> How reliable is scipy.sparse? Is there any plan to make it more
> compatible with the array type? Behave like the array type? How can I
> help?

I find it to be quite reliable for my work (i.e. WorksForMe), but I've 
been using spmat.matvec() since the beginning. I use SVN versions of 
scipy/numpy since the sparse bits seem to improve faster than the release 
cycles come around. Of course you can always help by reporting bugs and 
making patches for things you think could be better.

There's a lot to be done. The trick is not to be shy, if your patches are 
OK they'll get applied quickly, and the worst that can happen is that 
someone will say no ;P

Regards
Neilen



From gael.varoquaux at normalesup.org  Fri Nov  2 10:08:21 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Fri, 2 Nov 2007 15:08:21 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
Message-ID: <20071102140821.GA1517@clipper.ens.fr>

On Wed, Oct 31, 2007 at 11:51:07PM +0100, Matthieu Brucher wrote:
>    2007/10/31, Paul Barrett <[1]pebarrett at gmail.com>:

>      No, you are not dreaming.  I also commented on this possibility
>      several years ago to some people in the SciPy community.
>      Unfortunately, I have not had the time or motivation to act on it.  It
>      would certainly be nice to have a basic package to show people.
>      You'll need to create a set of APIs so that you can connect to
>      instrumentation.  It shouldn't be too hard now that the interfaces
>      have been standardized.

>    Sorry, it was in fact Vision and Viper I saw in the past as well.
>    The idea is great but the application, IMHO, is not quite optimal :
>    - even for the tutorial, the rotate bloc should have two inputs, the image
>    and the rotation angle (this last factor could be given by the result of
>    another algorithm)
>    - what about multithreading in this model ?
>    - what about flow control ? (which is really a huge issue that is lacking
>    in ITK for instance)
>    - is there a way to create a bunch of blocks easily ?

>    All this has consequences on the interface. Besides, it would be very
>    great to use Traits as input and output blocks.

Enthought has a very promising deveolppement in this direction, that uses
inspection of a Python script to build the diagram. IMHO it is much more
powerful than vision because most of the work is done automaticaly. Look
at Eric Jones' presentation at this year's Scipy to read a bit more about
it.

Vision is really impressive. I am just worried that it is hard to reuse.
It is an empire of its own. 

As a general note on the mailing list, there are a lot of impressive
projects, but often they are centered on themselves, and do not provide a
good interface and reusable components. Recently Prabhu and I have been
working on Mayavi2 to make it more reusable, turning it into an engine
and widget that can be decoupled from the UI. This kind of approach is
very important to avoid duplication of effort. Concerning vision, there
are at least 3 or 4 projects doing Visual programming for scientific
Python purposes. It tears my heart to see that many different efforts not
sharing code, but the reason is simply that each project is hard to
reuse.

An, yes, I agree that traits would do a perfect basis to build a reusable
visual progamming model. Traits do need a better packaging and
deployement, but that is a problem much easier to solve than build a new
GUI dialog engine.

My 2cents,

Gael


From unpingco.jose at gmail.com  Fri Nov  2 10:11:14 2007
From: unpingco.jose at gmail.com (Jose Unpingco)
Date: Fri, 2 Nov 2007 07:11:14 -0700
Subject: [SciPy-user] communications toolbox for scientific python?
Message-ID: <b2967d1f0711020711x175488e7t52839d2f97557e7b@mail.gmail.com>

Is there anything like the Matlab communications toolbox in scientific
python?

Otherwise, is there anybody out there interested in putting one together?


Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071102/9cc6e3e1/attachment.html>

From matthieu.brucher at gmail.com  Fri Nov  2 10:11:58 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 2 Nov 2007 15:11:58 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <20071102140821.GA1517@clipper.ens.fr>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr>
Message-ID: <e76aa17f0711020711vad2b777p78c3dd7fd2a6f339@mail.gmail.com>

>
> Enthought has a very promising deveolppement in this direction, that uses
> inspection of a Python script to build the diagram. IMHO it is much more
> powerful than vision because most of the work is done automaticaly. Look
> at Eric Jones' presentation at this year's Scipy to read a bit more about
> it.



Do you have a link to the slides perhaps ?


As a general note on the mailing list, there are a lot of impressive
> projects, but often they are centered on themselves, and do not provide a
> good interface and reusable components. Recently Prabhu and I have been
> working on Mayavi2 to make it more reusable, turning it into an engine
> and widget that can be decoupled from the UI. This kind of approach is
> very important to avoid duplication of effort. Concerning vision, there
> are at least 3 or 4 projects doing Visual programming for scientific
> Python purposes. It tears my heart to see that many different efforts not
> sharing code, but the reason is simply that each project is hard to
> reuse.



Exactly. This means a common way of describing the blocks/plugins, a common
library to interact wth the user -> traits ?, ...


An, yes, I agree that traits would do a perfect basis to build a reusable
> visual progamming model. Traits do need a better packaging and
> deployement, but that is a problem much easier to solve than build a new
> GUI dialog engine.
>
> My 2cents,
>
> Gael
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071102/a65272ae/attachment.html>

From s.mientki at ru.nl  Fri Nov  2 14:31:53 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Fri, 02 Nov 2007 19:31:53 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <20071102140821.GA1517@clipper.ens.fr>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr>
Message-ID: <472B6D19.50409@ru.nl>

hi Gael,


Gael Varoquaux wrote:
> On Wed, Oct 31, 2007 at 11:51:07PM +0100, Matthieu Brucher wrote:
>   
>>    2007/10/31, Paul Barrett <[1]pebarrett at gmail.com>:
>>     
>
>   
>>      No, you are not dreaming.  I also commented on this possibility
>>      several years ago to some people in the SciPy community.
>>      Unfortunately, I have not had the time or motivation to act on it.  It
>>      would certainly be nice to have a basic package to show people.
>>      You'll need to create a set of APIs so that you can connect to
>>      instrumentation.  It shouldn't be too hard now that the interfaces
>>      have been standardized.
>>     
>
>   
>>    Sorry, it was in fact Vision and Viper I saw in the past as well.
>>    The idea is great but the application, IMHO, is not quite optimal :
>>    - even for the tutorial, the rotate bloc should have two inputs, the image
>>    and the rotation angle (this last factor could be given by the result of
>>    another algorithm)
>>    - what about multithreading in this model ?
>>    - what about flow control ? (which is really a huge issue that is lacking
>>    in ITK for instance)
>>    - is there a way to create a bunch of blocks easily ?
>>     
>
>   
>>    All this has consequences on the interface. Besides, it would be very
>>    great to use Traits as input and output blocks.
>>     
>
> Enthought has a very promising deveolppement in this direction, that uses
> inspection of a Python script to build the diagram. IMHO it is much more
> powerful than vision because most of the work is done automaticaly. Look
> at Eric Jones' presentation at this year's Scipy to read a bit more about
> it.
>
> Vision is really impressive. I am just worried that it is hard to reuse.
> It is an empire of its own. 
>
> As a general note on the mailing list, there are a lot of impressive
> projects, but often they are centered on themselves, and do not provide a
> good interface and reusable components. Recently Prabhu and I have been
> working on Mayavi2 to make it more reusable, turning it into an engine
> and widget that can be decoupled from the UI. This kind of approach is
> very important to avoid duplication of effort. Concerning vision, there
> are at least 3 or 4 projects doing Visual programming for scientific
> Python purposes.
Well I'm very interested to see these projects !
>  It tears my heart to see that many different efforts not
> sharing code, but the reason is simply that each project is hard to
> reuse.
>
> An, yes, I agree that traits would do a perfect basis to build a reusable
> visual progamming model. Traits do need a better packaging and
> deployement, but that is a problem much easier to solve than build a new
> GUI dialog engine.
>
> My 2cents,
>
> Gael
>   
I'm not sure how to react on this message,
I think Enthought has done an incredible job,
in making the Scipy Enthought Edition, ...
... and probably Traits is remarkable development, but ...
... I don't understand one bit of it !!
So it's either much too high brow for me, or it's brought too much high 
brow.
I can't find a few simple examples that I can download, with the 
guarantee that my current Scipy environment is untouched.
I took a very quick look at one of your powerpoints presentations, more 
than 95% text ??
are we taking about a UI ??

Gael, forget my words,
make it more open and  easier to step into.

cheers,
Stef Mientki






From anand.prabhakar.patil at gmail.com  Fri Nov  2 20:03:25 2007
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Fri, 2 Nov 2007 17:03:25 -0700
Subject: [SciPy-user] Memory quirk
Message-ID: <2bc7a5a50711021703n6d5e138fq8c10124dec711b77@mail.gmail.com>

Hi all,

The following script leaks memory when an array with dtype=object is used
instead of a list. Is this a bug, or a behavior I should be expecting? Any
help is much appreciated.

Thanks,
Anand


from numpy import *
import time
import gc

class MyObject(object):
    def __init__(self):
        self.data = zeros(100000, dtype=float)

        # This leaks memory...
        self.stufflist = array([self], dtype=object)

        # This doesn't
        # self.stufflist = [self]

for i in xrange(10000):
    q=MyObject()
    time.sleep(.01)
    gc.collect()
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071102/cf47a0e8/attachment.html>

From anand.prabhakar.patil at gmail.com  Fri Nov  2 21:40:23 2007
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Fri, 2 Nov 2007 18:40:23 -0700
Subject: [SciPy-user] Another memory quirk
Message-ID: <2bc7a5a50711021840r29580f44o793624072f2e2ef3@mail.gmail.com>

Hi all,

Not sure if this is related to my last post but the script below leaks
memory like mad... PLEASE DON'T RUN IT, I don't want to be responsible for
crashing your computer!

Is there any safe way to store a ravelled view of an array in a subclass
like this? I'm using the ravelled view inside a very tight loop & would
rather not spend time calling ravel() if possible.

Many thanks in advance,
Anand


from numpy import *
from PyMC import *

class MyArray(ndarray):
    pass

for i in xrange(1000):

    A=zeros((1000,1000),dtype=float)

    A = A.view(MyArray)

    # If I uncomment this and run it, a huge memory leak ensues.
    # A._ravelled = A.ravel()
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071102/ad846aec/attachment.html>

From oliphant at enthought.com  Sat Nov  3 02:55:35 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Sat, 03 Nov 2007 01:55:35 -0500
Subject: [SciPy-user] [Fwd: Re:  Memory quirk]
Message-ID: <472C1B67.1000703@enthought.com>


-------------- next part --------------
An embedded message was scrubbed...
From: "Travis E. Oliphant" <oliphant at enthought.com>
Subject: Re: [SciPy-user] Memory quirk
Date: Sat, 03 Nov 2007 01:13:32 -0500
Size: 1391
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071103/3212b64c/attachment.eml>

From oliphant at enthought.com  Sat Nov  3 02:57:49 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Sat, 03 Nov 2007 01:57:49 -0500
Subject: [SciPy-user] Another memory quirk
In-Reply-To: <2bc7a5a50711021840r29580f44o793624072f2e2ef3@mail.gmail.com>
References: <2bc7a5a50711021840r29580f44o793624072f2e2ef3@mail.gmail.com>
Message-ID: <472C1BED.4060904@enthought.com>

Anand Patil wrote:
> Hi all,
>
> Not sure if this is related to my last post but the script below leaks 
> memory like mad... PLEASE DON'T RUN IT, I don't want to be responsible 
> for crashing your computer!
>
> Is there any safe way to store a ravelled view of an array in a 
> subclass like this? I'm using the ravelled view inside a very tight 
> loop & would rather not spend time calling ravel() if possible.
>
> Many thanks in advance,
> Anand

Here you have created another cyclic reference.

A.ravel() returns a new array object which references the old one (in 
the .base attribute).  In this case you are storing this new array 
object into the ._ravelled attribute of the old one.  Thus, the old 
MyArray object (the one returned from the view statement) is being 
referenced indirectly by-itself through its "_ravelled" attribute.   The 
reference count never goes to zero on the MyArray object and these 
cycles just blossom off each holding on to their original  ~8 MB array.

You've just shown two good examples where object cycles using arrays can 
hurt you.   It's an important point.   It may be possible to add cyclic 
garbage collection to NumPy objects (and it's sub-classes),  but it 
won't be trivial).

In this case, just keep the original array reference and set the 
_ravelled attribute to a raveled version of the original array:

B = zeros((1000,1000), float)
A = B.view(MyArray)
A._ravelled = B.ravel()

This way A is not referencing itself through one of its attributes and 
will be deleted once A gets re-assigned.

-Travis









>
>
> from numpy import *
> from PyMC import *
>    
> class MyArray(ndarray):
>     pass
>
> for i in xrange(1000):
>    
>     A=zeros((1000,1000),dtype=float)
>    
>     A = A.view(MyArray)
>
>     # If I uncomment this and run it, a huge memory leak ensues.
>     # A._ravelled = A.ravel()
>    
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From gael.varoquaux at normalesup.org  Sat Nov  3 10:25:35 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sat, 3 Nov 2007 15:25:35 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <e76aa17f0711020711vad2b777p78c3dd7fd2a6f339@mail.gmail.com>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr>
	<e76aa17f0711020711vad2b777p78c3dd7fd2a6f339@mail.gmail.com>
Message-ID: <20071103142535.GD1956@clipper.ens.fr>

On Fri, Nov 02, 2007 at 03:11:58PM +0100, Matthieu Brucher wrote:
>      Look at Eric Jones' presentation at this year's Scipy to read a
>      bit more about it.

>    Do you have a link to the slides perhaps ?

Sorry, I was on the road, with very poor internet access, so I wrote the
e-mail a bit quickly.

The link is http://scipy.org/SciPy2007/TutorialMaterials.

Ga?l


From gael.varoquaux at normalesup.org  Sat Nov  3 10:52:33 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sat, 3 Nov 2007 15:52:33 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472B6D19.50409@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
Message-ID: <20071103145233.GE1956@clipper.ens.fr>

On Fri, Nov 02, 2007 at 07:31:53PM +0100, Stef Mientki wrote:
> > Concerning vision, there are at least 3 or 4 projects doing Visual
> > programming for scientific Python purposes.

> Well I'm very interested to see these projects !

OK, now I have to name them, huho. 

So we have Vision, that is indeed impressive, but really built around
custom tools. It would be a large amount of work to abstract out the
visual programming part of the rest, but it is probably possible.

Then we have orange, that dies no seem to be fully written in Python, but
certainly uses Python (http://www.ailab.si/orange/screenshots.asp)

Elefant also uses a visual programming engine. see
http://scipy.org/SciPy2007/Presentations for Christfried Webers' slides
on elefant, or their main webpage (
http://elefant.developer.nicta.com.au/ ). I talked to Christfried at
scipy, and he definitly seemed open for collaborations.

And we have enthought's approach. If you want to have a glimpse at the
UI, you can look on page 24 of Eric's slides. The two interesting things
about enthought's approach is that they reuse their knowhow in layout
engine to generate the dialogs (I am not to sure it is pure traits, but it
seems linked), and, most interesting, that they use standard python
functions to build their visual blocks, and do not need descriptions of
input and output. Obviously this imposes a lot of restriction on the way
the functions behave (basically pure functionnal-like behavior, where
side effects are banned), and last time I talked with them, they had
problems adding flow control to this. But they are using this engine in a
commercial app, which means that there is money going in it to improve
it, so it is definetaly worth the look.

Ga?l


From nmarais at sun.ac.za  Sat Nov  3 11:50:42 2007
From: nmarais at sun.ac.za (Neilen Marais)
Date: Sat, 3 Nov 2007 15:50:42 +0000 (UTC)
Subject: [SciPy-user] communications toolbox for scientific python?
References: <b2967d1f0711020711x175488e7t52839d2f97557e7b@mail.gmail.com>
Message-ID: <fgi5ci$uji$2@ger.gmane.org>

Hi,

On Fri, 02 Nov 2007 07:11:14 -0700, Jose Unpingco wrote:

> Is there anything like the Matlab communications toolbox in scientific
> python?
> 
> Otherwise, is there anybody out there interested in putting one
> together?

I'm not personally familiar with what the Matlab comms toolbox can do. I 
think a good contribution would be to describe the functionality that 
exists in that toolbox in a way that doesn't reference Matlab. The scipy 
wiki would no doubt be a good place to do that. It would then be much 
easier for individual bits to get implemented. I might even find that 
some of those bits overlap with my needs, giving my selfish little self 
incentive to work on them :)

Regards
Neilen

Regards
Neilen



> http://projects.scipy.org/mailman/listinfo/scipy-user




From anand.prabhakar.patil at gmail.com  Sat Nov  3 13:31:20 2007
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Sat, 3 Nov 2007 10:31:20 -0700
Subject: [SciPy-user] Memory quirk
Message-ID: <2bc7a5a50711031031m486818bua66d34eb2fdea3c3@mail.gmail.com>

Hi Travis, thanks for your help. I think I get it but your response to my
first posting didn't come through, would you mind resending?

Also, if possible it would be nice to have this information in Guide to
NumPy. I had no idea numpy did its own garbage collecting, so I had no clue
where to start looking.

Cheers
Anand

------------------------------
>
> Message: 4
> Date: Sat, 03 Nov 2007 01:55:35 -0500
> From: "Travis E. Oliphant" <oliphant at enthought.com>
> Subject: [SciPy-user] [Fwd: Re:  Memory quirk]
> To: SciPy Users List <scipy-user at scipy.org>
> Message-ID: <472C1B67.1000703 at enthought.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071103/1eb29656/attachment.html>

From oliphant at enthought.com  Sat Nov  3 15:52:05 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Sat, 03 Nov 2007 14:52:05 -0500
Subject: [SciPy-user] Memory quirk
In-Reply-To: <2bc7a5a50711021703n6d5e138fq8c10124dec711b77@mail.gmail.com>
References: <2bc7a5a50711021703n6d5e138fq8c10124dec711b77@mail.gmail.com>
Message-ID: <472CD165.8000800@enthought.com>

Anand Patil wrote:
> Hi all,
>
> The following script leaks memory when an array with dtype=object is 
> used instead of a list. Is this a bug, or a behavior I should be 
> expecting? Any help is much appreciated.
It's not really a bug, just an unsupported use of NumPy object arrays
which do not participate in Python's cyclic garbage collector.    To add
support would require more re-factoring than it has been worth to
anybody at this point.

It would require handing object array allocations differently so that
they participate in the garbage collection strategy and it may not be
possible (I'm not sure if an object can sometimes have it's memory
allocated using it's own allocators and sometimes have them allocated
using PyObject_GC_New and still participate in the garbage collection
cycle).

-Travis




From s.mientki at ru.nl  Sat Nov  3 18:52:52 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Sat, 03 Nov 2007 23:52:52 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <20071103145233.GE1956@clipper.ens.fr>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr>
Message-ID: <472CFBC4.1090802@ru.nl>



Gael Varoquaux wrote:
> On Fri, Nov 02, 2007 at 07:31:53PM +0100, Stef Mientki wrote:
>   
>>> Concerning vision, there are at least 3 or 4 projects doing Visual
>>> programming for scientific Python purposes.
>>>       
>
>   
>> Well I'm very interested to see these projects !
>>     
>
> OK, now I have to name them, huho.
thanks very much Gael,
it's very good to have such an overview.
>  
>
> So we have Vision, that is indeed impressive, but really built around
> custom tools. It would be a large amount of work to abstract out the
> visual programming part of the rest, but it is probably possible.
>   
Yes I took a quick look, and doesn't seems easy to extract the Vision 
part form the total package.
> Then we have orange, that dies no seem to be fully written in Python, but
> certainly uses Python (http://www.ailab.si/orange/screenshots.asp)
>
>   
Looks nice, but is very dedicated to statistics  (and Qt-based :-(  ,
so I don't know if we can build a hardware control loop in it.
If you get more library modules, it will be hard to manage them.
> Elefant also uses a visual programming engine. see
> http://scipy.org/SciPy2007/Presentations for Christfried Webers' slides
> on elefant, or their main webpage (
> http://elefant.developer.nicta.com.au/ ). I talked to Christfried at
> scipy, and he definitly seemed open for collaborations.
>   
Looks very nice, also statistical oriented, graphing is almost the same 
as ogl-like.py.
Not yet suited for windows :-(
> And we have enthought's approach. If you want to have a glimpse at the
> UI, you can look on page 24 of Eric's slides.
couldn't find that :-(
>  The two interesting things
> about enthought's approach is that they reuse their knowhow in layout
> engine to generate the dialogs (I am not to sure it is pure traits, but it
> seems linked), and, most interesting, that they use standard python
> functions to build their visual blocks, and do not need descriptions of
> input and output. Obviously this imposes a lot of restriction on the way
> the functions behave (basically pure functionnal-like behavior, where
> side effects are banned), and last time I talked with them, they had
> problems adding flow control to this. But they are using this engine in a
> commercial app, which means that there is money going in it to improve
> it, so it is definetaly worth the look.
>   
I don't think any of the mentioned packages is capable of (pseudo-) 
realtime programming,
so there might be room for one more package ;-)

cheers,
Stef



From gael.varoquaux at normalesup.org  Sat Nov  3 19:07:01 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 4 Nov 2007 00:07:01 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472CFBC4.1090802@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
Message-ID: <20071103230701.GL6725@clipper.ens.fr>

On Sat, Nov 03, 2007 at 11:52:52PM +0100, Stef Mientki wrote:
> I don't think any of the mentioned packages is capable of (pseudo-) 
> realtime programming,

What do you mean by that. Reaction to the changes as they happen, a bit
like in labview. I think enthought's pacakage would be the one to look
at, as a reactive interaction with the user was one of their design
goals.

> so there might be room for one more package ;-)

No, you mean, it would be interesting to extend an existing package :->.

Ga?l


From s.mientki at ru.nl  Sat Nov  3 20:37:28 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Sun, 04 Nov 2007 01:37:28 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <20071103230701.GL6725@clipper.ens.fr>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
	<20071103230701.GL6725@clipper.ens.fr>
Message-ID: <472D1448.2090901@ru.nl>



Gael Varoquaux wrote:
> On Sat, Nov 03, 2007 at 11:52:52PM +0100, Stef Mientki wrote:
>   
>> I don't think any of the mentioned packages is capable of (pseudo-) 
>> realtime programming,
>>     
>
> What do you mean by that. Reaction to the changes as they happen, a bit
> like in labview. I think enthought's pacakage would be the one to look
> at, as a reactive interaction with the user was one of their design
> goals.
>   
What I mean, is for "realtime" control loops.
Let's say I sample data at something like 20 kHz,
want to perform some calculations on it,
display it and feed the manipulated data back to my hardware,
how am I going to let the data flow:
- sample by sample, much too slow
- chunking, that will work, but do these environments support that ?
I do the latter in Scipy, but with a Delphi GUI ;-) , that works quit well.
In the radio world there seems to be a even much faster approach (real 
time demodulation).
>> so there might be room for one more package ;-)
>>     
>
> No, you mean, it would be interesting to extend an existing package :->.
>   
It's always funny to see that "others" know better what "I" mean ;-)

Well Yes and No, seeing the mentioned examples, these are my quick 
thoughts (and it's not all about taste)
- I like the Elefant / SciCos graphical view
- I like the rectangle connections of Vision
- Libraries should be organized in a tree, which consists of an 
auto-organizing part and a user configurable part, this tree should have 
full drag and drop and be fully editable
- Dataflow should be selectable per block: sample-based, chunked or 
sequential
- Each block should be directly editable, which will automatically 
create a new device (unless ..)
- A block can be created by pure code, flow chart, graphical state 
machine or any combination of these,
but also from an electrical circuit diagram, a pneumatic circuit, a 
physical design or a virtual world design.
- In sequential mode, partial execution should be possible, which means 
that all intermediate results should be stored
- The user interface of the final application should be controlled 
completely, independent of the functional program
- Not the deep nesting of LabView, everything should be available in at 
most 2 clicks
- Code snippets should be available in user organizable way

ok, I'm dreaming again ;-)

but on the other hand, most of the parts are somewhere in Python available !

cheers,
Stef



From gael.varoquaux at normalesup.org  Sat Nov  3 20:52:13 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 4 Nov 2007 01:52:13 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472D1448.2090901@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
	<20071103230701.GL6725@clipper.ens.fr> <472D1448.2090901@ru.nl>
Message-ID: <20071104005213.GA17516@clipper.ens.fr>

On Sun, Nov 04, 2007 at 01:37:28AM +0100, Stef Mientki wrote:
> What I mean, is for "realtime" control loops.
> Let's say I sample data at something like 20 kHz,
> want to perform some calculations on it,
> display it and feed the manipulated data back to my hardware,
> how am I going to let the data flow:
> - sample by sample, much too slow
> - chunking, that will work, but do these environments support that ?
> I do the latter in Scipy, but with a Delphi GUI ;-) , that works quit well.
> In the radio world there seems to be a even much faster approach (real 
> time demodulation).

I have done this kind of stuff, and I don't think the current status of
the different Python visual programming engines is good enough to do
this, unless you want really low timescales for you feedback loop. I am
not saying it is impossible, I am just saying there is a lot of work to
be done. The problem is that unless your time scales are really slow, you
will want some reasonnably fine control on what is happening in you loop.

My 2 cents,

Ga?l


From mark at mitre.org  Sat Nov  3 21:08:45 2007
From: mark at mitre.org (Mark Heslep)
Date: Sat, 03 Nov 2007 21:08:45 -0400
Subject: [SciPy-user] nullclines for nonlinear ODEs
In-Reply-To: <88e473830710300926l66c350c3k72b7fda9e6fd5c31@mail.gmail.com>
References: <88e473830710300926l66c350c3k72b7fda9e6fd5c31@mail.gmail.com>
Message-ID: <472D1B9D.5050003@mitre.org>

John Hunter wrote:
> In a workshop on scientific computing in python  I ran with Andrew
> Straw last weekend at the Claremont Colleges, I presented a unit on
> ODEs, numerically integrating the Lotka-Volterra equations, drawing a
> direction field, plotting a trajectory in the  phase plane, and
> plotting the nullclines.  For the nullclines, I used matplotlib's
> contouring routine with levels=[0], but I was wondering if there was a
> better/more direct route.
>
>   
Anyone care to comment on scipy's useability for solving ODEs versus, 
say, Mathematica or Matlab?  I'm not very familiar w/ either of the 'M' 
packages, but may have them forced on me soon unless I'm able to say, 
no, we are better off w/ scipy for user metric 1, 2, ...  If scipy comes 
up short in some way, sign me up as able and willing to help.  Apologies 
for hijacking your thread, your example is perfect.

Mark


From mark at mitre.org  Sat Nov  3 22:34:08 2007
From: mark at mitre.org (Mark Heslep)
Date: Sat, 03 Nov 2007 22:34:08 -0400
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <fg59ep$47e$1@ger.gmane.org>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>	<p0623091bc34a7a6964ba@192.168.1.105>	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org>
Message-ID: <472D2FA0.3010001@mitre.org>


>
> The animation, all figures and all calculations were done in Python
> (+Scilab+Pylab), but obviously this is not a programming paper and Python
> is a research tool here (which I think is the goal of the creators of the
> wonderful language and libraries). Unfortunately there is only code for
> Matlab - just a few one-liners. In HRV (heart rate variability) research
> most people use Matlab and hardly anyone even knows about Python. I am
> trying to advertise Python, but the paper would not have been accepted with
> code in Python. 
>
> Of course the paper clearly says that the calculations were done in Python. 
>
> Jarek
>
>   
Given that you felt for the sake of publication you must use Matlab, was 
it worth it then to use scipy to begin with? That is, if you knew you 
would need to convert to Matlab, was scipy worth the trouble?


From evanlezar at gmail.com  Sun Nov  4 04:37:42 2007
From: evanlezar at gmail.com (Evan Lezar)
Date: Sun, 4 Nov 2007 11:37:42 +0200
Subject: [SciPy-user] resizing a dok_matrix
Message-ID: <dd61a9250711040137n7199c26er9d321c5a8a0d4fa4@mail.gmail.com>

Hi

I have been using a dok_matrix from scipy 0.5.3 to implement re-sizable
sparse matrices of which the final size is not known initially.  This was
done by adjusting the shape property of an object inherited from the
dok_matrix class.

My problem is that in scipy 0.6 the dok_matrix is no longer re-sizable in
this way.  Is there a specific reason for this?  Any suggestions will be
welcomed.

Thanks
Evan

-- 
visit http://randomestrandom.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/f6ccca5f/attachment.html>

From samrobertsmith at gmail.com  Sun Nov  4 05:08:21 2007
From: samrobertsmith at gmail.com (linda.s)
Date: Sun, 4 Nov 2007 02:08:21 -0800
Subject: [SciPy-user] about MLab
Message-ID: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com>

I have quite a few codes written in Numeric.
MLab comes with Numeric too.
I got confused when I began to use numpy because it does not have MLab
and does not have some same functions in Numeric.
is there any smooth way to change the code developed in Numeric to numpy?
Thanks,
Linda


From C.J.Lee at tnw.utwente.nl  Sun Nov  4 06:06:31 2007
From: C.J.Lee at tnw.utwente.nl (C.J.Lee at tnw.utwente.nl)
Date: Sun, 4 Nov 2007 12:06:31 +0100
Subject: [SciPy-user] about MLab
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com>
Message-ID: <9E7F7554631C0047AB598527A4E9569B6F8102@tnx4.dynamic.tnw.utwente.nl>

Some Numeric functionality can be obtained as follows:

The old Numeric routine names and behavior for functions in the add-on packages UserArray, Matrix, MLab, LinearAlgebra, FFT, RandomArray, RNG and RNG.Statistics have been moved to numpy.oldnumeric.<pkg> where <pkg> is user_array, matrix, mlab, linear_algebra, fft, random_array, rng, and rng_stats respectively

Hope this helps
Cheers
Chris


-----Original Message-----
From: scipy-user-bounces at scipy.org on behalf of linda.s
Sent: Sun 11/4/2007 11:08 AM
To: SciPy Users List
Subject: [SciPy-user] about MLab
 
I have quite a few codes written in Numeric.
MLab comes with Numeric too.
I got confused when I began to use numpy because it does not have MLab
and does not have some same functions in Numeric.
is there any smooth way to change the code developed in Numeric to numpy?
Thanks,
Linda
_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org
http://projects.scipy.org/mailman/listinfo/scipy-user



From samrobertsmith at gmail.com  Sun Nov  4 06:33:04 2007
From: samrobertsmith at gmail.com (linda.s)
Date: Sun, 4 Nov 2007 03:33:04 -0800
Subject: [SciPy-user] about MLab
In-Reply-To: <9E7F7554631C0047AB598527A4E9569B6F8102@tnx4.dynamic.tnw.utwente.nl>
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com>
	<9E7F7554631C0047AB598527A4E9569B6F8102@tnx4.dynamic.tnw.utwente.nl>
Message-ID: <1d987df30711040333w5aaa7fe1r49fa6ccd109d6f87@mail.gmail.com>

>>> numpy.oldnumeric.mlab
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'oldnumeric'


On Nov 4, 2007 3:06 AM,  <C.J.Lee at tnw.utwente.nl> wrote:
> Some Numeric functionality can be obtained as follows:
>
> The old Numeric routine names and behavior for functions in the add-on packages UserArray, Matrix, MLab, LinearAlgebra, FFT, RandomArray, RNG and RNG.Statistics have been moved to numpy.oldnumeric.<pkg> where <pkg> is user_array, matrix, mlab, linear_algebra, fft, random_array, rng, and rng_stats respectively
>
> Hope this helps
> Cheers
> Chris
>
>
>
> -----Original Message-----
> From: scipy-user-bounces at scipy.org on behalf of linda.s
> Sent: Sun 11/4/2007 11:08 AM
> To: SciPy Users List
> Subject: [SciPy-user] about MLab
>
> I have quite a few codes written in Numeric.
> MLab comes with Numeric too.
> I got confused when I began to use numpy because it does not have MLab
> and does not have some same functions in Numeric.
> is there any smooth way to change the code developed in Numeric to numpy?
> Thanks,
> Linda
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From jaropis at gazeta.pl  Sun Nov  4 06:50:36 2007
From: jaropis at gazeta.pl (jaropis)
Date: Sun, 04 Nov 2007 12:50:36 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>	<p0623091bc34a7a6964ba@192.168.1.105>	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
Message-ID: <fgkbko$8t1$1@ger.gmane.org>

> Given that you felt for the sake of publication you must use Matlab, was
> it worth it then to use scipy to begin with? That is, if you knew you
> would need to convert to Matlab, was scipy worth the trouble?

A very good question. I use scipy because my research is multidisciplinary
and multicenter. This means that I have to write code that will be used in
various places all around the world. I cannot expect all my friends to buy
Matlab. The second reason is that I do not own a copy of Matlab - I used
the copy owned by my University only to test the code. If I had wanted to
use Matlab for this paper I would have had to spend weeks at the
University. Using Python I could work at home. The third reason is probably
the most important: Python is a real programming language and therefore it
is much more convenient. I get medical time series in various formats,
usualy some kind of text files, and python is great at processing them and
unifying all the formats I come accross to the single format which is used
by my software. Everyone who has ever tried to process gigabytes of text
files with Matlab knows that this is extremally difficult. It is also much
easier to write small scripts that have a specific function and send them
over the net to those who need them. 

As I said - the use of Matlab was necessary to include the code for the
various descriptors introduced in the paper - unfortunately Scipy code
would not have been accepted (I know this for a fact). The Python software
used in this paper was much more complex than the Matlab one-liners - I had
to process the recordings and do the calculations. Also, I had to write a
speparate program to produce the animation. Of course this software was not
converted to Matlab - just the mathematical expressions. 

Jarek



From matthieu.brucher at gmail.com  Sun Nov  4 06:58:15 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 4 Nov 2007 12:58:15 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472D1448.2090901@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
	<20071103230701.GL6725@clipper.ens.fr> <472D1448.2090901@ru.nl>
Message-ID: <e76aa17f0711040358y3fdaffa7p6789c87837255c0a@mail.gmail.com>

>
> > What do you mean by that. Reaction to the changes as they happen, a bit
> > like in labview. I think enthought's pacakage would be the one to look
> > at, as a reactive interaction with the user was one of their design
> > goals.
> What I mean, is for "realtime" control loops.
> Let's say I sample data at something like 20 kHz,
> want to perform some calculations on it,
> display it and feed the manipulated data back to my hardware,
> how am I going to let the data flow:
> - sample by sample, much too slow
> - chunking, that will work, but do these environments support that ?
> I do the latter in Scipy, but with a Delphi GUI ;-) , that works quit
> well.
> In the radio world there seems to be a even much faster approach (real
> time demodulation).
>

This needs, IMHO, a special package. Usual algorithms are not realtime
capable.
Real-time arises when there is a time notion, and this needs the use of
chunks and thus a special interface. For instance in audio, the standard is
hard to use, and there are a lot of pitfalls when implementing a host that
respects the different latencies.

Matthieu

-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/3422b020/attachment.html>

From C.J.Lee at tnw.utwente.nl  Sun Nov  4 07:06:23 2007
From: C.J.Lee at tnw.utwente.nl (C.J.Lee at tnw.utwente.nl)
Date: Sun, 4 Nov 2007 13:06:23 +0100
Subject: [SciPy-user] about MLab
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com><9E7F7554631C0047AB598527A4E9569B6F8102@tnx4.dynamic.tnw.utwente.nl>
	<1d987df30711040333w5aaa7fe1r49fa6ccd109d6f87@mail.gmail.com>
Message-ID: <9E7F7554631C0047AB598527A4E9569B6F8103@tnx4.dynamic.tnw.utwente.nl>


I would guess that means you have an incomplete install of numpy.

Check that in the site-packages directory that numpy/oldnumeric/mlab.py exists. If it doesn't then you need to reinstall numpy

I can't check on my linux or windows machine right now, but it certainly works on my Mac OS X box

Cheers
Chris


-----Original Message-----
From: scipy-user-bounces at scipy.org on behalf of linda.s
Sent: Sun 11/4/2007 12:33 PM
To: SciPy Users List
Subject: Re: [SciPy-user] about MLab
 
>>> numpy.oldnumeric.mlab
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'oldnumeric'


On Nov 4, 2007 3:06 AM,  <C.J.Lee at tnw.utwente.nl> wrote:
> Some Numeric functionality can be obtained as follows:
>
> The old Numeric routine names and behavior for functions in the add-on packages UserArray, Matrix, MLab, LinearAlgebra, FFT, RandomArray, RNG and RNG.Statistics have been moved to numpy.oldnumeric.<pkg> where <pkg> is user_array, matrix, mlab, linear_algebra, fft, random_array, rng, and rng_stats respectively
>
> Hope this helps
> Cheers
> Chris
>
>
>
> -----Original Message-----
> From: scipy-user-bounces at scipy.org on behalf of linda.s
> Sent: Sun 11/4/2007 11:08 AM
> To: SciPy Users List
> Subject: [SciPy-user] about MLab
>
> I have quite a few codes written in Numeric.
> MLab comes with Numeric too.
> I got confused when I began to use numpy because it does not have MLab
> and does not have some same functions in Numeric.
> is there any smooth way to change the code developed in Numeric to numpy?
> Thanks,
> Linda
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org
http://projects.scipy.org/mailman/listinfo/scipy-user



From matthieu.brucher at gmail.com  Sun Nov  4 07:57:33 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 4 Nov 2007 13:57:33 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472D1448.2090901@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
	<20071103230701.GL6725@clipper.ens.fr> <472D1448.2090901@ru.nl>
Message-ID: <e76aa17f0711040457q2143f08fp5b111d09a2fe05de@mail.gmail.com>

>
> - A block can be created by pure code, flow chart, graphical state
> machine or any combination of these,
> but also from an electrical circuit diagram, a pneumatic circuit, a
> physical design or a virtual world design.
>

The Simulink twin in Python ?

-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/c7f385ef/attachment.html>

From matthieu.brucher at gmail.com  Sun Nov  4 08:01:02 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 4 Nov 2007 14:01:02 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <fgkbko$8t1$1@ger.gmane.org>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
Message-ID: <e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>

>
> As I said - the use of Matlab was necessary to include the code for the
> various descriptors introduced in the paper - unfortunately Scipy code
> would not have been accepted (I know this for a fact). The Python software
> used in this paper was much more complex than the Matlab one-liners - I
> had
> to process the recordings and do the calculations. Also, I had to write a
> speparate program to produce the animation. Of course this software was
> not
> converted to Matlab - just the mathematical expressions.
>

In my field, the same problem arises with a Matlab toolbox, SPM. You can't
write an article without using this toolbox, especially if you want to
publish it in NeuroImage (the editor in chief is one of the writter of the
toolbox).

Matthieu
-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/83a7506c/attachment.html>

From et.gaudrain at free.fr  Sun Nov  4 08:03:13 2007
From: et.gaudrain at free.fr (Etienne Gaudrain)
Date: Sun, 04 Nov 2007 14:03:13 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <e76aa17f0711040358y3fdaffa7p6789c87837255c0a@mail.gmail.com>
References: <4728FFC8.2070007@ru.nl>	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>	<20071102140821.GA1517@clipper.ens.fr>
	<472B6D19.50409@ru.nl>	<20071103145233.GE1956@clipper.ens.fr>
	<472CFBC4.1090802@ru.nl>	<20071103230701.GL6725@clipper.ens.fr>
	<472D1448.2090901@ru.nl>
	<e76aa17f0711040358y3fdaffa7p6789c87837255c0a@mail.gmail.com>
Message-ID: <472DC311.40606@free.fr>

Hi,

I'm a Python newbie, and I'm not particularly familiar with LabView...

However, I know a soft/language that could be interesting to mention in 
this discussion.

The FAUST language is a functional language for audio processing. The 
language describes a block diagram (that can be generated in SVG from 
the code). One particularity is that the "FAUST compiler" reduces the 
process (using lambda-calculus) and then translates it in C++. In 
addition, the C++ code can then be wrapped with a GUI wrapper (GTK, 
wx...) and a sound system wrapper (ALSA, OSS...). The goal is cleary to 
interact with users during realtime processing.

I thought a few months ago that it could be just wonderful to be able to 
use FAUST code in Python just like done with Weave.
The developers of FAUST have developed an online version so you can test 
it here: http://faust.grame.fr/.

-Etienne



Matthieu Brucher a ?crit :
>
>     > What do you mean by that. Reaction to the changes as they
>     happen, a bit
>     > like in labview. I think enthought's pacakage would be the one
>     to look
>     > at, as a reactive interaction with the user was one of their design
>     > goals.
>     What I mean, is for "realtime" control loops.
>     Let's say I sample data at something like 20 kHz,
>     want to perform some calculations on it,
>     display it and feed the manipulated data back to my hardware,
>     how am I going to let the data flow:
>     - sample by sample, much too slow
>     - chunking, that will work, but do these environments support that ?
>     I do the latter in Scipy, but with a Delphi GUI ;-) , that works
>     quit well.
>     In the radio world there seems to be a even much faster approach (real
>     time demodulation).
>
>
> This needs, IMHO, a special package. Usual algorithms are not realtime 
> capable.
> Real-time arises when there is a time notion, and this needs the use 
> of chunks and thus a special interface. For instance in audio, the 
> standard is hard to use, and there are a lot of pitfalls when 
> implementing a host that respects the different latencies.
>
> Matthieu
>
> -- 
> French PhD student
> Website : http://miles.developpez.com/
> Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
> LinkedIn : http://www.linkedin.com/in/matthieubrucher
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   


-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Etienne Gaudrain
Universite Claude Bernard LYON 1
CNRS - UMR5020, Neurosciences Sensorielles, Comportement, Cognition
50, avenue Tony Garnier
69366 LYON Cedex 07
FRANCE
T?l : 04 37 28 74 85
Fax : 04 37 28 76 01
Page web equipe : http://olfac.univ-lyon1.fr/unite/equipe-02/equipe-2-f.html
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/e9da4067/attachment.html>

From aisaac at american.edu  Sun Nov  4 08:12:34 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Sun, 4 Nov 2007 08:12:34 -0500
Subject: [SciPy-user] about MLab
In-Reply-To: <1d987df30711040333w5aaa7fe1r49fa6ccd109d6f87@mail.gmail.com>
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com><9E7F7554631C0047AB598527A4E9569B6F8102@tnx4.dynamic.tnw.utwente.nl><1d987df30711040333w5aaa7fe1r49fa6ccd109d6f87@mail.gmail.com>
Message-ID: <Mahogany-0.67.0-1808-20071104-081234.00@american.edu>

On Sun, 4 Nov 2007, "linda.s" apparently wrote:
> >>> numpy.oldnumeric.mlab 
>  Traceback (most recent call last):
>    File "<stdin>", line 1, in <module>
>  AttributeError: 'module' object has no attribute 
>  'oldnumeric'



>>> from numpy.oldnumeric import mlab

hth,
Alan Isaac





From matthieu.brucher at gmail.com  Sun Nov  4 08:23:15 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 4 Nov 2007 14:23:15 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472CFBC4.1090802@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
Message-ID: <e76aa17f0711040523v31942facw58c2f61611ff051e@mail.gmail.com>

>
> > Elefant also uses a visual programming engine. see
> > http://scipy.org/SciPy2007/Presentations for Christfried Webers' slides
> > on elefant, or their main webpage (
> > http://elefant.developer.nicta.com.au/ ). I talked to Christfried at
> > scipy, and he definitly seemed open for collaborations.
> >
> Looks very nice, also statistical oriented, graphing is almost the same
> as ogl-like.py.
> Not yet suited for windows :-(
>

Why is that ? If this is because no binaries are available for Windows, this
can be done.

Matthieu
-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/7cbe42a2/attachment.html>

From s.mientki at ru.nl  Sun Nov  4 09:31:24 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Sun, 04 Nov 2007 15:31:24 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <e76aa17f0711040457q2143f08fp5b111d09a2fe05de@mail.gmail.com>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
	<20071103230701.GL6725@clipper.ens.fr> <472D1448.2090901@ru.nl>
	<e76aa17f0711040457q2143f08fp5b111d09a2fe05de@mail.gmail.com>
Message-ID: <472DD7BC.4070104@ru.nl>



Matthieu Brucher wrote:
>
>     - A block can be created by pure code, flow chart, graphical state
>     machine or any combination of these,
>     but also from an electrical circuit diagram, a pneumatic circuit, a
>     physical design or a virtual world design.
>
>
> The Simulink twin in Python ?
Simulinh AND PowerSim ;-)
Stef



From s.mientki at ru.nl  Sun Nov  4 10:10:21 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Sun, 04 Nov 2007 16:10:21 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <e76aa17f0711040523v31942facw58c2f61611ff051e@mail.gmail.com>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
	<e76aa17f0711040523v31942facw58c2f61611ff051e@mail.gmail.com>
Message-ID: <472DE0DD.6090900@ru.nl>



Matthieu Brucher wrote:
>
>     > Elefant also uses a visual programming engine. see
>     > http://scipy.org/SciPy2007/Presentations for Christfried Webers'
>     slides
>     > on elefant, or their main webpage (
>     > http://elefant.developer.nicta.com.au/ ). I talked to Christfried at
>     > scipy, and he definitly seemed open for collaborations.
>     >
>     Looks very nice, also statistical oriented, graphing is almost the
>     same
>     as ogl-like.py.
>     Not yet suited for windows :-(
>
>
> Why is that ? If this is because no binaries are available for 
> Windows, this can be done.
Probably yes, but not within my capabilities ;-)
cheers,
Stef



From unpingco.jose at gmail.com  Sun Nov  4 10:13:06 2007
From: unpingco.jose at gmail.com (Jose Unpingco)
Date: Sun, 4 Nov 2007 07:13:06 -0800
Subject: [SciPy-user] Scipy + Vision = LabView
Message-ID: <b2967d1f0711040713u66f99094u904904e01dee2d16@mail.gmail.com>

All:

I put the tutorial together, and I plan to do more along these lines.
My interest is in using VISION as  a framework for a detailed
communications simulation, although up to this point, VISION has been
primarily used as a tool in chemistry and biology.

I plan to write certain communications and signal processing
components based upon scipy and integrate them into VISION as a
distinct block set (i.e. component library). The developers of VISION
are situated nearby and I am looking forward to that collaboration.

Progress will be announced on the VISION mailing list
(http://mgldev.scripps.edu/mailman/listinfo/vision) and on my blog
(http://www.osc.edu/blogs/index.php/sip), but not on this list to
avoid polluting it.

As always,  I appreciate your criticism.

--Jos?


From s.mientki at ru.nl  Sun Nov  4 10:20:22 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Sun, 04 Nov 2007 16:20:22 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
Message-ID: <472DE336.5060007@ru.nl>



Matthieu Brucher wrote:
>
>     As I said - the use of Matlab was necessary to include the code
>     for the
>     various descriptors introduced in the paper - unfortunately Scipy
>     code
>     would not have been accepted (I know this for a fact). The Python
>     software
>     used in this paper was much more complex than the Matlab
>     one-liners - I had
>     to process the recordings and do the calculations. Also, I had to
>     write a
>     speparate program to produce the animation. Of course this
>     software was not
>     converted to Matlab - just the mathematical expressions.
>
>
> In my field, the same problem arises with a Matlab toolbox, SPM. You 
> can't write an article without using this toolbox, especially if you 
> want to publish it in NeuroImage (the editor in chief is one of the 
> writter of the toolbox).
This is really SAD:
you're doing medical research, investigating and describing some 
physiological phenomena,
well I can assume they are interested what algorithm you used,
to verify your conclusions,
but it's absolutely absurd, they make it obligatory that you write your 
manuscript with a pen of a certain brand !!

It's even more absurd, if you realize
that in MatLab the algorithms are unknown,
so you can't verify them in case of weird results.
( I had a few a couple of years ago,
and the answer of MathWorks was "try to get a job at MathWorks" ;-)

But rethinking this attitude, maybe it's not absurd at all:
because there are unknown errors in MatLab,
they are unable to verify your results if you did the analysis in SciPy 
;-) ;-)

cheers,
Stef Mientki





From matthew.brett at gmail.com  Sun Nov  4 10:39:49 2007
From: matthew.brett at gmail.com (Matthew Brett)
Date: Sun, 4 Nov 2007 07:39:49 -0800
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <472DE336.5060007@ru.nl>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
	<472DE336.5060007@ru.nl>
Message-ID: <1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>

Hi,

> > In my field, the same problem arises with a Matlab toolbox, SPM. You
> > can't write an article without using this toolbox, especially if you
> > want to publish it in NeuroImage (the editor in chief is one of the
> > writter of the toolbox).
> This is really SAD:
> you're doing medical research, investigating and describing some
> physiological phenomena,
> well I can assume they are interested what algorithm you used,
> to verify your conclusions,
> but it's absolutely absurd, they make it obligatory that you write your
> manuscript with a pen of a certain brand !!

Hmm - I'm in the same field, and trying to work out a way to fix this with:

http://neuroimaging.scipy.org

The problem is not as silly as it seems.  Neuroimaging is moderately
complex, and it's difficult for reviewers to understand what has been
done if you have used a variety of software packages.  The advantage
of SPM, in this case, is that the code for the neuroimaging part is
open source (GPL), it covers the whole range of processing (more or
less), and it's fairly easy to extend.  The disadvantage is that - you
need matlab - and that matlab is so clunky as a programming language
that it deals very badly with the increasing need for complexity,
meaning it is harder and harder to develop in SPM as time goes by.
Our thought was that the only way out of this was to develop a system
that would allow you to call the core algorithms of SPM within python,
and develop new ones that you could validate against SPM (and other
tools) as the standard.  But it's hard work getting that right,
because our community is so fragmented.

Best,

Matthew

Matthew


From et.gaudrain at free.fr  Sun Nov  4 10:57:29 2007
From: et.gaudrain at free.fr (Etienne Gaudrain)
Date: Sun, 04 Nov 2007 16:57:29 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <472DE336.5060007@ru.nl>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>	<p0623091bc34a7a6964ba@192.168.1.105>	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>	<fg59ep$47e$1@ger.gmane.org>
	<472D2FA0.3010001@mitre.org>	<fgkbko$8t1$1@ger.gmane.org>	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
	<472DE336.5060007@ru.nl>
Message-ID: <472DEBE9.4030905@free.fr>

Hi,

I understand you consider this is absurd. In a way, it is. And I agree 
that it is not pleasant to have to go back to Matlab when you moved up 
to Python.

However, scientific publication have to meet a few standards and beside 
the infightings and abuse of power of some editors, we have to consider 
that scientific publication:
- must be archived and archives must be reachable
- must be readable/understandable by the community
- must describe things that one could reproduce

The two last points are, for now, not good for Python. Most of the 
community still don't know that Python exists. Then they consider 
themselves as unable to understand Python code (even if anybody who read 
in Matlab can actually read in Python). Worth, they don't have Python 
installed on their computer, so they can't run the .py. Ok, this no big 
deal, since Python can be freely downloaded. But, publications don't 
come along with a tutorial on how to install Python... (and Numpy.... 
and Scipy... and Pylab... and...)

Finally, what is the most important in choosing between Python and 
Matlab is that Matlab is far more broadly used is the community than 
Python. For now. In scientific publication, the most common language 
must be used, even if it is not optimal. Look, we (try to) write papers 
in English, while French is far more elegant :-P .

So, since Matlab is the standard, editors will require source codes in 
Matlab. But The Mathworks has recently done some mistakes in Matlab 
distributions (broke some backward compatibility for example), and a 
part of the community is moving to Python. So we can hope that since 
Scipy developement look stable, and since the documentation of 
Python/Numpy/Scipy is reliableenough, the community will largely move to 
Python. And will make substantial economies of money and time :-D .

Regards,
-Etienne



Stef Mientki a ?crit :
> Matthieu Brucher wrote:
>   
>>     As I said - the use of Matlab was necessary to include the code
>>     for the
>>     various descriptors introduced in the paper - unfortunately Scipy
>>     code
>>     would not have been accepted (I know this for a fact). The Python
>>     software
>>     used in this paper was much more complex than the Matlab
>>     one-liners - I had
>>     to process the recordings and do the calculations. Also, I had to
>>     write a
>>     speparate program to produce the animation. Of course this
>>     software was not
>>     converted to Matlab - just the mathematical expressions.
>>
>>
>> In my field, the same problem arises with a Matlab toolbox, SPM. You 
>> can't write an article without using this toolbox, especially if you 
>> want to publish it in NeuroImage (the editor in chief is one of the 
>> writter of the toolbox).
>>     
> This is really SAD:
> you're doing medical research, investigating and describing some 
> physiological phenomena,
> well I can assume they are interested what algorithm you used,
> to verify your conclusions,
> but it's absolutely absurd, they make it obligatory that you write your 
> manuscript with a pen of a certain brand !!
>
> It's even more absurd, if you realize
> that in MatLab the algorithms are unknown,
> so you can't verify them in case of weird results.
> ( I had a few a couple of years ago,
> and the answer of MathWorks was "try to get a job at MathWorks" ;-)
>
> But rethinking this attitude, maybe it's not absurd at all:
> because there are unknown errors in MatLab,
> they are unable to verify your results if you did the analysis in SciPy 
> ;-) ;-)
>
> cheers,
> Stef Mientki
>
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>   


-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Etienne Gaudrain
Universite Claude Bernard LYON 1
CNRS - UMR5020, Neurosciences Sensorielles, Comportement, Cognition
50, avenue Tony Garnier
69366 LYON Cedex 07
FRANCE
T?l : 04 37 28 74 85
Fax : 04 37 28 76 01
Page web equipe : http://olfac.univ-lyon1.fr/unite/equipe-02/equipe-2-f.html
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/74aaccea/attachment.html>

From et.gaudrain at free.fr  Sun Nov  4 11:02:49 2007
From: et.gaudrain at free.fr (Etienne Gaudrain)
Date: Sun, 04 Nov 2007 17:02:49 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>	<p0623091bc34a7a6964ba@192.168.1.105>	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>	<fg59ep$47e$1@ger.gmane.org>
	<472D2FA0.3010001@mitre.org>	<fgkbko$8t1$1@ger.gmane.org>	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>	<472DE336.5060007@ru.nl>
	<1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
Message-ID: <472DED29.1030101@free.fr>

Is there any attempt to merge this project with http://neuralensemble.org/ ?

-Etienne



Matthew Brett a ?crit :
> Hi,
>
> Hmm - I'm in the same field, and trying to work out a way to fix this with:
>
> http://neuroimaging.scipy.org
>
> The problem is not as silly as it seems.  Neuroimaging is moderately
> complex, and it's difficult for reviewers to understand what has been
> done if you have used a variety of software packages.  The advantage
> of SPM, in this case, is that the code for the neuroimaging part is
> open source (GPL), it covers the whole range of processing (more or
> less), and it's fairly easy to extend.  The disadvantage is that - you
> need matlab - and that matlab is so clunky as a programming language
> that it deals very badly with the increasing need for complexity,
> meaning it is harder and harder to develop in SPM as time goes by.
> Our thought was that the only way out of this was to develop a system
> that would allow you to call the core algorithms of SPM within python,
> and develop new ones that you could validate against SPM (and other
> tools) as the standard.  But it's hard work getting that right,
> because our community is so fragmented.
>
> Best,
>
> Matthew
>
> Matthew
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>   


-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Etienne Gaudrain
Universite Claude Bernard LYON 1
CNRS - UMR5020, Neurosciences Sensorielles, Comportement, Cognition
50, avenue Tony Garnier
69366 LYON Cedex 07
FRANCE
T?l : 04 37 28 74 85
Fax : 04 37 28 76 01
Page web equipe : http://olfac.univ-lyon1.fr/unite/equipe-02/equipe-2-f.html
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 



From nicolas.pettiaux at ael.be  Sun Nov  4 11:09:57 2007
From: nicolas.pettiaux at ael.be (Nicolas Pettiaux)
Date: Sun, 4 Nov 2007 17:09:57 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <fg59ep$47e$1@ger.gmane.org>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org>
Message-ID: <f89206d80711040809vee2682agf137ef96fc7fd30c@mail.gmail.com>

2007/10/29, jaropis <jaropis at gazeta.pl>:
> > Maybe Jarek could send us or post on his website a copy of a previous
> > version of the paper. These are not covered usually by the (abusive)
> > copyright of the Journal
>
> No problem - the copyright agreement gives me the right to post it on my
> website, so here it is:
>
> http://www.if.uz.zgora.pl/~jaropis/geomasy

much thanks,

Nicolas

> The animation, all figures and all calculations were done in Python
> (+Scilab+Pylab), but obviously this is not a programming paper and Python
> is a research tool here (which I think is the goal of the creators of the
> wonderful language and libraries). Unfortunately there is only code for
> Matlab - just a few one-liners. In HRV (heart rate variability) research
> most people use Matlab and hardly anyone even knows about Python. I am
> trying to advertise Python, but the paper would not have been accepted with
> code in Python.

such a dependance on the language is very sad. As well as is the fact
IMHO, that in a scientific paper, the author do not have to publish
the code, or at least allow the other researchers to have access to it
and read it and rerun it, to check the results as this one of the very
purpose of publishing.

Nicolas
-- 
Nicolas Pettiaux - email: nicolas.pettiaux at ael.be
Utiliser des formats ouverts et des logiciels libres -
http://www.passeralinux.org.
Pour la bureautique, les seuls formats ISO sont ceux de http://fr.openoffice.org


From matthew.brett at gmail.com  Sun Nov  4 11:41:36 2007
From: matthew.brett at gmail.com (Matthew Brett)
Date: Sun, 4 Nov 2007 08:41:36 -0800
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <472DED29.1030101@free.fr>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
	<472DE336.5060007@ru.nl>
	<1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
	<472DED29.1030101@free.fr>
Message-ID: <1e2af89e0711040841h362f1fa1v94340e28a0707fa0@mail.gmail.com>

Hi,

On 11/4/07, Etienne Gaudrain <et.gaudrain at free.fr> wrote:
> Is there any attempt to merge this project with http://neuralensemble.org/ ?

No, not so far - but neural network modelling and neuroimaging are
very different areas, and there's very little overlap in the
algorithms used.  Neuroimaging is mainly image processing for example.

Best,

Matthew


From oliphant at enthought.com  Sun Nov  4 11:48:09 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Sun, 04 Nov 2007 10:48:09 -0600
Subject: [SciPy-user] about MLab
In-Reply-To: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com>
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com>
Message-ID: <472DF7C9.8020305@enthought.com>

linda.s wrote:
> I have quite a few codes written in Numeric.
> MLab comes with Numeric too.
> I got confused when I began to use numpy because it does not have MLab
> and does not have some same functions in Numeric.
> is there any smooth way to change the code developed in Numeric to numpy?
>   

You should also read the first chapters of the book "Guide to NumPy" 
which are available on-line for immediate download.  There is a 
comprehensive description of the differences between Numeric and NumPy 
as well as tips for compatibility.

You must also import numpy.oldnumeric before it is available.

import numpy.oldnumeric

numpy.oldnumeric.mlab

You will notice that the module is simple a re-naming module.  All of 
the functionality is in numpy but under possibly different names.

All functions in Numeric are available in NumPy (as far as I know).  If 
you encounter a situation where that is not true, please let me know.

-Travis



From matthieu.brucher at gmail.com  Sun Nov  4 11:48:07 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 4 Nov 2007 17:48:07 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472DE0DD.6090900@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
	<e76aa17f0711040523v31942facw58c2f61611ff051e@mail.gmail.com>
	<472DE0DD.6090900@ru.nl>
Message-ID: <e76aa17f0711040848t79d61a74r43e12ed329e66b0c@mail.gmail.com>

I'm searching a way of using the GUI, but I didn't find it ATM, even for
Linux.
I'm noticed that it used its own "OGL", this explains the better look and
feel than Viper and OGL-based applications.

Matthieu

> Why is that ? If this is because no binaries are available for
> > Windows, this can be done.
> Probably yes, but not within my capabilities ;-)
> cheers,
> Stef
>

Matthieu
-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/782f8e15/attachment.html>

From matthieu.brucher at gmail.com  Sun Nov  4 11:52:59 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 4 Nov 2007 17:52:59 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
	<472DE336.5060007@ru.nl>
	<1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
Message-ID: <e76aa17f0711040852h5d2282cfn922b4aff0c0a15a@mail.gmail.com>

Hi,

It's even worse. If you design a new algorithm with results that cannot be
verified with SPM, you can't ublish anything (worse : the neurologists do
not want to publish results with your algorithms).
The fact that SPM is GPL is not relevent for me. What bothers me is that
there is a conflict of interest between the SPM team and the editorial board
of some journals et the fact that even you demonstrated that your algorithms
are correct, you can't publish something that SPM cannot coroborate (knowing
that SPM relies on some heavy hypothesis that are not even correct).

Matthieu

PS : I really hope nipy will go on growing and that people at my lab will
collaborate, at last, even if it is not for mainstream algorithms (SPMs,
...)

2007/11/4, Matthew Brett <matthew.brett at gmail.com>:
>
> Hi,
>
> > > In my field, the same problem arises with a Matlab toolbox, SPM. You
> > > can't write an article without using this toolbox, especially if you
> > > want to publish it in NeuroImage (the editor in chief is one of the
> > > writter of the toolbox).
> > This is really SAD:
> > you're doing medical research, investigating and describing some
> > physiological phenomena,
> > well I can assume they are interested what algorithm you used,
> > to verify your conclusions,
> > but it's absolutely absurd, they make it obligatory that you write your
> > manuscript with a pen of a certain brand !!
>
> Hmm - I'm in the same field, and trying to work out a way to fix this
> with:
>
> http://neuroimaging.scipy.org
>
> The problem is not as silly as it seems.  Neuroimaging is moderately
> complex, and it's difficult for reviewers to understand what has been
> done if you have used a variety of software packages.  The advantage
> of SPM, in this case, is that the code for the neuroimaging part is
> open source (GPL), it covers the whole range of processing (more or
> less), and it's fairly easy to extend.  The disadvantage is that - you
> need matlab - and that matlab is so clunky as a programming language
> that it deals very badly with the increasing need for complexity,
> meaning it is harder and harder to develop in SPM as time goes by.
> Our thought was that the only way out of this was to develop a system
> that would allow you to call the core algorithms of SPM within python,
> and develop new ones that you could validate against SPM (and other
> tools) as the standard.  But it's hard work getting that right,
> because our community is so fragmented.
>
> Best,
>
> Matthew
>
> Matthew
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/b9f4da0f/attachment.html>

From jh at physics.ucf.edu  Sun Nov  4 11:53:01 2007
From: jh at physics.ucf.edu (Joe Harrington)
Date: Sun, 04 Nov 2007 11:53:01 -0500
Subject: [SciPy-user] surface plotting
Message-ID: <1194195181.5621.364.camel@glup.physics.ucf.edu>

I'm in search of either Python or an open-source, standalone image
viewer that can visualize surface plots.  A surface plot is a rendering
of a 2D image as a 3D surface using the image values as the Z
coordinate.  The new wiremesh in matplotlib is way too slow to be of
practical use.  It takes over a minute to plot a 1024**2 image, and for
my purposes I need to move the surface around and look at it from
different angles, which requires many re-renderings.

Previously, I had used IDL, which has two routines called surface and
shade_surface.  The first is a wiremesh, the second is the same but adds
shading according to viewing angle.  These routines ran at decent speeds
in IDL even when they were introduced about 20 years ago.  IRAF also has
them (or at least the wiremesh), but its size and klunkiness make it a
poor solution as well.  I'm not sure what's up with matplotlib, but this
is a very new feature and it just might not be doing its thing in the
most efficient way.  In the long run, perhaps the source code for the
IRAF routine could be a guide for doing it efficiently.

In the short run, any ideas?  It would make a good cookbook recipe.

--jh--




From matthieu.brucher at gmail.com  Sun Nov  4 11:54:50 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 4 Nov 2007 17:54:50 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <472DEBE9.4030905@free.fr>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
	<472DE336.5060007@ru.nl> <472DEBE9.4030905@free.fr>
Message-ID: <e76aa17f0711040854g100dff74o2640d764d52ed2b0@mail.gmail.com>

Well, and in some communities, if your work is done in Matlab and not in C
or C++, your work is not even looked at... (not even talking about the fact
that generally you don't even say if your code is in X or Y)

Matthieu

2007/11/4, Etienne Gaudrain <et.gaudrain at free.fr>:
>
>  Hi,
>
> I understand you consider this is absurd. In a way, it is. And I agree
> that it is not pleasant to have to go back to Matlab when you moved up to
> Python.
>
> However, scientific publication have to meet a few standards and beside
> the infightings and abuse of power of some editors, we have to consider that
> scientific publication:
> - must be archived and archives must be reachable
> - must be readable/understandable by the community
> - must describe things that one could reproduce
>
> The two last points are, for now, not good for Python. Most of the
> community still don't know that Python exists. Then they consider themselves
> as unable to understand Python code (even if anybody who read in Matlab can
> actually read in Python). Worth, they don't have Python installed on their
> computer, so they can't run the .py. Ok, this no big deal, since Python can
> be freely downloaded. But, publications don't come along with a tutorial on
> how to install Python... (and Numpy.... and Scipy... and Pylab... and...)
>
> Finally, what is the most important in choosing between Python and Matlab
> is that Matlab is far more broadly used is the community than Python. For
> now. In scientific publication, the most common language must be used, even
> if it is not optimal. Look, we (try to) write papers in English, while
> French is far more elegant :-P .
>
> So, since Matlab is the standard, editors will require source codes in
> Matlab. But The Mathworks has recently done some mistakes in Matlab
> distributions (broke some backward compatibility for example), and a part of
> the community is moving to Python. So we can hope that since Scipy
> developement look stable, and since the documentation of Python/Numpy/Scipy
> is reliableenough, the community will largely move to Python. And will make
> substantial economies of money and time :-D .
>
> Regards,
> -Etienne
>
>
>
> Stef Mientki a ?crit :
>
> Matthieu Brucher wrote:
>
>      As I said - the use of Matlab was necessary to include the code
>     for the
>     various descriptors introduced in the paper - unfortunately Scipy
>     code
>     would not have been accepted (I know this for a fact). The Python
>     software
>     used in this paper was much more complex than the Matlab
>     one-liners - I had
>     to process the recordings and do the calculations. Also, I had to
>     write a
>     speparate program to produce the animation. Of course this
>     software was not
>     converted to Matlab - just the mathematical expressions.
>
>
> In my field, the same problem arises with a Matlab toolbox, SPM. You
> can't write an article without using this toolbox, especially if you
> want to publish it in NeuroImage (the editor in chief is one of the
> writter of the toolbox).
>
>  This is really SAD:
> you're doing medical research, investigating and describing some
> physiological phenomena,
> well I can assume they are interested what algorithm you used,
> to verify your conclusions,
> but it's absolutely absurd, they make it obligatory that you write your
> manuscript with a pen of a certain brand !!
>
> It's even more absurd, if you realize
> that in MatLab the algorithms are unknown,
> so you can't verify them in case of weird results.
> ( I had a few a couple of years ago,
> and the answer of MathWorks was "try to get a job at MathWorks" ;-)
>
> But rethinking this attitude, maybe it's not absurd at all:
> because there are unknown errors in MatLab,
> they are unable to verify your results if you did the analysis in SciPy
> ;-) ;-)
>
> cheers,
> Stef Mientki
>
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Etienne Gaudrain
> Universite Claude Bernard LYON 1
> CNRS - UMR5020, Neurosciences Sensorielles, Comportement, Cognition
> 50, avenue Tony Garnier
> 69366 LYON Cedex 07
> FRANCE
> T?l : 04 37 28 74 85
> Fax : 04 37 28 76 01
> Page web equipe : http://olfac.univ-lyon1.fr/unite/equipe-02/equipe-2-f.html
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/752692fd/attachment.html>

From gael.varoquaux at normalesup.org  Sun Nov  4 12:19:04 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 4 Nov 2007 18:19:04 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <1194195181.5621.364.camel@glup.physics.ucf.edu>
References: <1194195181.5621.364.camel@glup.physics.ucf.edu>
Message-ID: <20071104171904.GB29007@clipper.ens.fr>

On Sun, Nov 04, 2007 at 11:53:01AM -0500, Joe Harrington wrote:
> I'm in search of either Python or an open-source, standalone image
> viewer that can visualize surface plots.  A surface plot is a rendering
> of a 2D image as a 3D surface using the image values as the Z
> coordinate.  

Mayavi2 can do this, and has a scripting interface that should be easy to
use for what you want to do. You can have a look at
https://svn.enthought.com/enthought/attachment/wiki/MayaVi/Sciy2007_mlab_slides.pdf?format=raw
for instance.

Hopefully you should find all the info you need to install mayavi2 on
https://svn.enthought.com/enthought/wiki/MayaVi, if not just ask.

HTH,

Ga?l


From matthew.brett at gmail.com  Sun Nov  4 12:21:54 2007
From: matthew.brett at gmail.com (Matthew Brett)
Date: Sun, 4 Nov 2007 17:21:54 +0000
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <e76aa17f0711040852h5d2282cfn922b4aff0c0a15a@mail.gmail.com>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<p0623091bc34a7a6964ba@192.168.1.105>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
	<472DE336.5060007@ru.nl>
	<1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
	<e76aa17f0711040852h5d2282cfn922b4aff0c0a15a@mail.gmail.com>
Message-ID: <1e2af89e0711040921vb19c7cdq9126ab1b6dd0162f@mail.gmail.com>

Hi,

> It's even worse. If you design a new algorithm with results that cannot be
> verified with SPM, you can't ublish anything (worse : the neurologists do
> not want to publish results with your algorithms).

Yes - I know what you mean.  What we are trying to do is collect some
support across the imaging community for NiPy as a valid alternative.
When it's functional (ahem).  Several of the team are well known in
the imaging community, so I think, if we produce something functional,
and publish on it, it will be hard to argue against using it for
published work.   Especially if you can show that using SPM (via
wrappers) gives the same (similar) result.  The problem is momentum.
We're very close to critical mass now, but somehow we haven't
succeeded in pooling our efforts enough to really break the back of
the work,

Matthew


From C.J.Lee at tnw.utwente.nl  Sun Nov  4 12:28:12 2007
From: C.J.Lee at tnw.utwente.nl (C.J.Lee at tnw.utwente.nl)
Date: Sun, 4 Nov 2007 18:28:12 +0100
Subject: [SciPy-user] Subject: Re: Scientific Python publications
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org><p0623091bc34a7a6964ba@192.168.1.105><f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com><fg59ep$47e$1@ger.gmane.org>
	<472D2FA0.3010001@mitre.org><fgkbko$8t1$1@ger.gmane.org><e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com><472DE336.5060007@ru.nl><1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
	<e76aa17f0711040852h5d2282cfn922b4aff0c0a15a@mail.gmail.com>
Message-ID: <9E7F7554631C0047AB598527A4E9569B6F8104@tnx4.dynamic.tnw.utwente.nl>

Perhaps a compromise(for the moment) is to provide both. So minimal matlab scripting to allow publishable and the full python code also available. That way the phd students learn that there is an alternative available and the matlab *only* attitude will die out with retirements. This way you meet the requirements for the peer review and you can grow the python neuroimaging community

It might also be interesting to attempt publishing results which show the differences in functionality between the matlab based code and python based code, particularly if you can find a test case where the matlab toolbox is definitively wrong. (I am not involved in neuroimaging so I can't really say anything about the feasibility of this)

Cheers
Chris


-----Original Message-----
From: scipy-user-bounces at scipy.org on behalf of Matthieu Brucher
Sent: Sun 11/4/2007 5:52 PM
To: SciPy Users List
Subject: Re: [SciPy-user] Subject: Re: Scientific Python publications
 
Hi,

It's even worse. If you design a new algorithm with results that cannot be
verified with SPM, you can't ublish anything (worse : the neurologists do
not want to publish results with your algorithms).
The fact that SPM is GPL is not relevent for me. What bothers me is that
there is a conflict of interest between the SPM team and the editorial board
of some journals et the fact that even you demonstrated that your algorithms
are correct, you can't publish something that SPM cannot coroborate (knowing
that SPM relies on some heavy hypothesis that are not even correct).

Matthieu

PS : I really hope nipy will go on growing and that people at my lab will
collaborate, at last, even if it is not for mainstream algorithms (SPMs,
...)

2007/11/4, Matthew Brett <matthew.brett at gmail.com>:
>
> Hi,
>
> > > In my field, the same problem arises with a Matlab toolbox, SPM. You
> > > can't write an article without using this toolbox, especially if you
> > > want to publish it in NeuroImage (the editor in chief is one of the
> > > writter of the toolbox).
> > This is really SAD:
> > you're doing medical research, investigating and describing some
> > physiological phenomena,
> > well I can assume they are interested what algorithm you used,
> > to verify your conclusions,
> > but it's absolutely absurd, they make it obligatory that you write your
> > manuscript with a pen of a certain brand !!
>
> Hmm - I'm in the same field, and trying to work out a way to fix this
> with:
>
> http://neuroimaging.scipy.org
>
> The problem is not as silly as it seems.  Neuroimaging is moderately
> complex, and it's difficult for reviewers to understand what has been
> done if you have used a variety of software packages.  The advantage
> of SPM, in this case, is that the code for the neuroimaging part is
> open source (GPL), it covers the whole range of processing (more or
> less), and it's fairly easy to extend.  The disadvantage is that - you
> need matlab - and that matlab is so clunky as a programming language
> that it deals very badly with the increasing need for complexity,
> meaning it is harder and harder to develop in SPM as time goes by.
> Our thought was that the only way out of this was to develop a system
> that would allow you to call the core algorithms of SPM within python,
> and develop new ones that you could validate against SPM (and other
> tools) as the standard.  But it's hard work getting that right,
> because our community is so fragmented.
>
> Best,
>
> Matthew
>
> Matthew
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher



From matthew.brett at gmail.com  Sun Nov  4 13:53:22 2007
From: matthew.brett at gmail.com (Matthew Brett)
Date: Sun, 4 Nov 2007 18:53:22 +0000
Subject: [SciPy-user] Subject: Re: Scientific Python publications
In-Reply-To: <9E7F7554631C0047AB598527A4E9569B6F8104@tnx4.dynamic.tnw.utwente.nl>
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org>
	<f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com>
	<fg59ep$47e$1@ger.gmane.org> <472D2FA0.3010001@mitre.org>
	<fgkbko$8t1$1@ger.gmane.org>
	<e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com>
	<472DE336.5060007@ru.nl>
	<1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com>
	<e76aa17f0711040852h5d2282cfn922b4aff0c0a15a@mail.gmail.com>
	<9E7F7554631C0047AB598527A4E9569B6F8104@tnx4.dynamic.tnw.utwente.nl>
Message-ID: <1e2af89e0711041053j4262a1b5t829a2003793110c3@mail.gmail.com>

Hi,

On 11/4/07, C.J.Lee at tnw.utwente.nl <C.J.Lee at tnw.utwente.nl> wrote:
> Perhaps a compromise(for the moment) is to provide both. So minimal matlab scripting to allow publishable and the full python code also available. That way the phd students learn that there is an alternative available and the matlab *only* attitude will die out with retirements. This way you meet the requirements for the peer review and you can grow the python neuroimaging community

Yes, that was more or less exactly our pitch to the NIH in the grant
that got funded recently.

> It might also be interesting to attempt publishing results which show the differences in functionality between the matlab based code and python based code, particularly if you can find a test case where the matlab toolbox is definitively wrong. (I am not involved in neuroimaging so I can't really say anything about the feasibility of this)

I should say, in fairness, that the algorithms in SPM are generally
very good, and I have found the main maintainer of SPM to be very
generous and open-minded.  For example, he wrote a very helpful and
thoughtful letter of recommendation for our grant.  For me, my main
problem with SPM is matlab, and the very great difficulty of building
a productive architecture with matlab.  I'm quite sure we can build
something much more fruitful in python, and would like to provide that
as a service to the community, including the SPM community, of which
I'm a long-time member.

Matthew


From karl.young at ucsf.edu  Sun Nov  4 14:23:47 2007
From: karl.young at ucsf.edu (Young, Karl)
Date: Sun, 4 Nov 2007 11:23:47 -0800
Subject: [SciPy-user] Subject: Re: Scientific Python publications
References: <mailman.104908.1192203753.30507.scipy-user@scipy.org><p0623091bc34a7a6964ba@192.168.1.105><f89206d80710281102o538b85efhe90b541a08e6618a@mail.gmail.com><fg59ep$47e$1@ger.gmane.org>
	<472D2FA0.3010001@mitre.org><fgkbko$8t1$1@ger.gmane.org><e76aa17f0711040501j28941c98w2a43cab1d601ed1c@mail.gmail.com><472DE336.5060007@ru.nl><1e2af89e0711040739i394bce23r4145f041bb2032d4@mail.gmail.com><472DED29.1030101@free.fr>
	<1e2af89e0711040841h362f1fa1v94340e28a0707fa0@mail.gmail.com>
Message-ID: <9D202D4E86A4BF47BA6943ABDF21BE78039F099B@EXVS06.net.ucsf.edu>



>> On 11/4/07, Etienne Gaudrain <et.gaudrain at free.fr> wrote:
>> Is there any attempt to merge this project with http://neuralensemble.org/ ?

> No, not so far - but neural network modelling and neuroimaging are
> very different areas, and there's very little overlap in the
> algorithms used.  Neuroimaging is mainly image processing for example.

Which is not to say that it might not be interesting at some future point. E.g. some recent modeling work by Sporns et. al. re. using a detailed map of the map of the Macaque brain to generate a dynamic network model that reproduced patterns seen in resting state fMRI is drawing these areas a little closer (as if there wasn't already enough on the plates of NiPy folks !).



From haase at msg.ucsf.edu  Sun Nov  4 15:22:14 2007
From: haase at msg.ucsf.edu (Sebastian Haase)
Date: Sun, 4 Nov 2007 21:22:14 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <20071104171904.GB29007@clipper.ens.fr>
References: <1194195181.5621.364.camel@glup.physics.ucf.edu>
	<20071104171904.GB29007@clipper.ens.fr>
Message-ID: <bc657ead0711041222o31df4dfx3e26282f2d28b692@mail.gmail.com>

On Nov 4, 2007 6:19 PM, Gael Varoquaux <gael.varoquaux at normalesup.org> wrote:
> On Sun, Nov 04, 2007 at 11:53:01AM -0500, Joe Harrington wrote:
> > I'm in search of either Python or an open-source, standalone image
> > viewer that can visualize surface plots.  A surface plot is a rendering
> > of a 2D image as a 3D surface using the image values as the Z
> > coordinate.
>
> Mayavi2 can do this, and has a scripting interface that should be easy to
> use for what you want to do. You can have a look at
> https://svn.enthought.com/enthought/attachment/wiki/MayaVi/Sciy2007_mlab_slides.pdf?format=raw
> for instance.
>
> Hopefully you should find all the info you need to install mayavi2 on
> https://svn.enthought.com/enthought/wiki/MayaVi, if not just ask.
>
Hi  Ga?l,
I'm following MayaVi for a long time.
And I'm looking forward to install the new (wx based version of)
MavaVi (mayavi2).

The page you referred to, however, lists the dependencies, which is
just daunting !!
VTK by itself is considered to be a difficult install... (you can't do
nothing about that).

But now, what is
TVTK ? (Does it depend on other parts of Enthough ?)
How do you get Traits, without having to get ALL the rest of Enthought ?
Last but not least, I'm not a big fan of setuptools: googling for it
turns up mostly scipy and some other projects - but no "official
endorsement" by python.org ! Some people warn that it might download
files and install them on your system without warning ....
In my particular case, I don't have system admin permissions, and it
seems to require (or at least, work best) when writing to the systems
site-packages folder.

Don't get me wrong: I'm not criticising, I'm very interested in most
of the above (especially Traits, BTW). So far I just could not manage
to start using any of them: neither on Linux (debian), nor on Windows.

Any help is very much appreciated.
-Sebastian Haase


From zunzun at zunzun.com  Sun Nov  4 16:44:24 2007
From: zunzun at zunzun.com (James Phillips)
Date: Sun, 4 Nov 2007 16:44:24 -0500
Subject: [SciPy-user] surface plotting
In-Reply-To: <1194195181.5621.364.camel@glup.physics.ucf.edu>
References: <1194195181.5621.364.camel@glup.physics.ucf.edu>
Message-ID: <268756d30711041344x17c2230y8d60a0b536e36e02@mail.gmail.com>

I use VTK to make VRML, which you can then rotate around
and view using a VRML viewer.  If you use Debian or Ubuntu,
a quick "sudo apt-get install python-vtk" will install Python
bindings and all dependencies.

http://www.vtk.org/

Here's a quick synopsis from IBM that's fairly recent, they
discuss MayaVi which also does what you're looking for - and
for debian/ubuntu, "sudo apt-get install mayavi":

http://www.ibm.com/developerworks/linux/library/l-datavistools/

Since MayaVi uses VTK, the examples use many of the VTK
examples:

http://mayavi.sourceforge.net/screenshots/index.html

     James



On 11/4/07, Joe Harrington <jh at physics.ucf.edu> wrote:
>
> I'm in search of either Python or an open-source, standalone image
> viewer that can visualize surface plots.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071104/0d787bb7/attachment.html>

From eike.welk at gmx.net  Sun Nov  4 17:13:27 2007
From: eike.welk at gmx.net (Eike Welk)
Date: Sun, 04 Nov 2007 23:13:27 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472D1448.2090901@ru.nl>
References: <4728FFC8.2070007@ru.nl> <20071103230701.GL6725@clipper.ens.fr>
	<472D1448.2090901@ru.nl>
Message-ID: <200711042313.28683.eike.welk@gmx.net>

On Sunday 04 November 2007 01:37, Stef Mientki wrote:
> What I mean, is for "realtime" control loops.
> Let's say I sample data at something like 20 kHz,
> want to perform some calculations on it,
> display it and feed the manipulated data back to my hardware,

For realtime and high speed there is also Gnuradio. 
http://www.gnu.org/software/gnuradio/

It is organized as a flow-graph, with the nodes written in C++. It 
works in the MHz range. The graph is constructed in Python and there 
is some facility to transfer results back to Python. It has no 
graphical front end.

I didn't try Gnuradio however. I read only some of the documentation:
http://www.gnu.org/software/gnuradio/doc/exploring-gnuradio.html#software

If a graphical front end would be written, a wrapper for Gnuradio and 
a method to transfer data to/from Numpy would be great.

Regards, Eike.


From gael.varoquaux at normalesup.org  Sun Nov  4 17:20:32 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 4 Nov 2007 23:20:32 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <bc657ead0711041222o31df4dfx3e26282f2d28b692@mail.gmail.com>
References: <1194195181.5621.364.camel@glup.physics.ucf.edu>
	<20071104171904.GB29007@clipper.ens.fr>
	<bc657ead0711041222o31df4dfx3e26282f2d28b692@mail.gmail.com>
Message-ID: <20071104222032.GD29007@clipper.ens.fr>

On Sun, Nov 04, 2007 at 09:22:14PM +0100, Sebastian Haase wrote:
> The page you referred to, however, lists the dependencies, which is
> just daunting !!
> VTK by itself is considered to be a difficult install... (you can't do
> nothing about that).

I agree. this is quite bad. We are working on a Debian package, so this
will settle the problem for Debian and Ubuntu.


> But now, what is
> TVTK ? (Does it depend on other parts of Enthough ?)
> How do you get Traits, without having to get ALL the rest of Enthought ?
> Last but not least, I'm not a big fan of setuptools: googling for it
> turns up mostly scipy and some other projects - but no "official
> endorsement" by python.org ! Some people warn that it might download
> files and install them on your system without warning ....

I agree completely with you. My trust in setuptools has gone down the
more I have used it.

You can install mayavi by getting the three tarballs on
http://code.enthought.com/downloads/source/ets2.6/, and building them.
The install is just a question of copying the resulting build dirs in
your PYTHONPATH, so that is simple. You might not have all the
dependencies for the build. I think the easiest is to to install the
Debian packages if you are under debian. They are listed on the install
page. If you are not under Debian (or alike), just state your system, and
we might be able to help you along a bit more.

> In my particular case, I don't have system admin permissions, and it
> seems to require (or at least, work best) when writing to the systems
> site-packages folder.

setuptools works very well when you don't have admin permissions, but it
doesn't solve the fact that you need vtk and swig installed. You can give
it a switch to tell it to write to another folder than the systems
site-packages.

> Don't get me wrong: I'm not criticising, I'm very interested in most
> of the above (especially Traits, BTW). So far I just could not manage
> to start using any of them: neither on Linux (debian), nor on Windows.

Under windows the simplest way to install mayavi in to install Python
using enstaller: http://code.enthought.com/enstaller/ , under Debian we
are working on Debian packages.

On a more general note, I have only recently realized that there was a
packaging problem. We are slowly working on improving that, but after I
finish my thesis (end of the year), I'll start some heavier work on
improving this, in particular providing Debians packages, and hopefully
binaries for other operating systems. Currently I am under deadline
pressure.

If you use the tarballs on
http://code.enthought.com/downloads/source/ets2.6/ you will need all the
dependencies, but no internet access, and building is separated from
installing. If you can only get the run-time dependencies, please state
precisely your system, and we can try building a binary tarball for you.
You will need:

VTK >= 4.4 
numpy >= 1.01
wxPython 2.6.x 

all the rest is either compile-time dependencies, or included in the bigs
tarballs.

HTH,

Ga?l


From s.mientki at ru.nl  Sun Nov  4 17:47:13 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Sun, 04 Nov 2007 23:47:13 +0100
Subject: [SciPy-user] [ANN] Scope_Plot,
	another plot library for real time signals.
Message-ID: <472E4BF1.2050905@ru.nl>

hello,

I justed finished, another plot library, called Scope_Plot, based on 
wxPython.

Scope_Plot is special meant for displaying real time signals,
and therefor has some new functionalities:
- signal selection
- each signal has it's own scale,
- moving erase block
- measurement cursor
ans should be a lot faster than MatPlot, Plot and FloatCanvas,
at least for real time signals (because it only draws the changes).

An animated demo can be seen here (2 MB, 2:10):
  http://stef.mientki.googlepages.com/jalspy_scope.html

A description of the library, as used in an application, can be found here:
  http://oase.uci.kun.nl/~mientki/data_www/pic/jalspy/jalspy_scope.html

And finally the file (and a few necessary libs) can be found here:
  http://oase.uci.kun.nl/~mientki/download/Scope_Plot.zip
The library has a main section which contains a simple demo,
the animated demo and the application description shows a more complex 
signal organization.

cheers,
Stef Mientki


From fredmfp at gmail.com  Sun Nov  4 19:10:45 2007
From: fredmfp at gmail.com (fred)
Date: Mon, 05 Nov 2007 01:10:45 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <bc657ead0711041222o31df4dfx3e26282f2d28b692@mail.gmail.com>
References: <1194195181.5621.364.camel@glup.physics.ucf.edu>	<20071104171904.GB29007@clipper.ens.fr>	<bc657ead0711041222o31df4dfx3e26282f2d28b692@mail.gmail.com>
Message-ID: <472E5F85.8050806@gmail.com>

On Sun, Nov 04, 2007 at 09:22:14PM +0100, Sebastian Haase wrote:
> The page you referred to, however, lists the dependencies, which is
> just daunting !!
> VTK by itself is considered to be a difficult install... (you can't do
> nothing about that).
What do you mean here ?

Which difficulties did you encounter ?

 From tarball sources ? from your packages manager ?

On which platforms ?

-- 
http://scipy.org/FredericPetit


From prabhu at aero.iitb.ac.in  Sun Nov  4 23:58:17 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Mon, 05 Nov 2007 10:28:17 +0530
Subject: [SciPy-user] surface plotting
In-Reply-To: <20071104222032.GD29007@clipper.ens.fr>
References: <1194195181.5621.364.camel@glup.physics.ucf.edu>	<20071104171904.GB29007@clipper.ens.fr>	<bc657ead0711041222o31df4dfx3e26282f2d28b692@mail.gmail.com>
	<20071104222032.GD29007@clipper.ens.fr>
Message-ID: <472EA2E9.7000206@aero.iitb.ac.in>

Gael Varoquaux wrote:
> On Sun, Nov 04, 2007 at 09:22:14PM +0100, Sebastian Haase wrote:
>> The page you referred to, however, lists the dependencies, which is
>> just daunting !!
>> VTK by itself is considered to be a difficult install... (you can't do
>> nothing about that).
> 
> I agree. this is quite bad. We are working on a Debian package, so this
> will settle the problem for Debian and Ubuntu.

VTK is packaged for Debian and on FC7.

> I agree completely with you. My trust in setuptools has gone down the
> more I have used it.

I'm afraid I don't quite share that sentiment as regards being a user of 
setuptools/easy_install.  Too often the version of setuptools installed 
in a distribution is old with bugs.  Right now if you are on ubuntu, 
feisty/gutsy or Debian it is almost trivial to install things using the 
binary eggs.

As Fred suggested, the best place for help on installation issues is the 
enthought-dev list.

cheers,
prabhu


From timmichelsen at gmx-topmail.de  Mon Nov  5 08:19:56 2007
From: timmichelsen at gmx-topmail.de (Timmie)
Date: Mon, 5 Nov 2007 13:19:56 +0000 (UTC)
Subject: [SciPy-user] time series analysis
Message-ID: <loom.20071105T131924-81@post.gmane.org>

Dear list members,
I am looking for some hints and recommendations.

I want to use python to analyse and evaluate (measurement) time series.

Therefore I am looking for some modules and documentation or tutorials which
could help me in

* filtering time series for a cetain criteria
* reduce them to a lower time resolution: e.g. from houly values to daily mean
values
* filling data gaps for days or hours without valid data using statistical
methods like regression
* general data quality and trend assessment

I found a conversion of the |Stat ("pipe-stat") programs written by Gary Perlman
into a python module: pstat.py -
http://www.nmr.mgh.harvard.edu/Neural_Systems_Group/gary/python/pstat.py

This can be used as a start. But I am interested in what you could recommend.

Thanks.

Kind regards,
Timmie






From gael.varoquaux at normalesup.org  Mon Nov  5 10:50:40 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 5 Nov 2007 16:50:40 +0100
Subject: [SciPy-user] Smoothing of a signal
In-Reply-To: <353163.97628.qm@web52306.mail.re2.yahoo.com>
References: <353163.97628.qm@web52306.mail.re2.yahoo.com>
Message-ID: <20071105155040.GE21399@clipper.ens.fr>

Hi,

On Mon, Nov 05, 2007 at 07:20:35AM -0800, Marc Beck wrote:
>    I am a graduate assistant at the Space Science Center at Morehead State
>    University. I am currently writing a program to convert raw data from our
>    21m antenna into a *.FITS image using numpy and scipy. I need to convolve
>    the pixels of my image. I have three arrays: one for the x- coordinate,
>    one for the y- coordinate and one for the brightness of the pixel. I need
>    to feed those three arrays into a function to convolve the pixels and I
>    need to get the same three arrays out with the new values, or the rest of
>    the program does not work. Someone from the python forum referred me to
>    some code you wrote at [1]http://www.scipy.org/Cookbook/SignalSmooth

>    but I don't understand which variable is what and how to integrate this
>    into my program. Please explain to me how I can make my program work. I
>    hope that I have described clearly enough what I am planning to do.

Your are better off asking the mailing list for these kind of question. I
am terribly busy currently (important deadlines) and almost deleted you
message before realising it had been sent only to me.

Is you data on a regular grid ? You currently have an (X, Y, V) data
layout, and it would be much easier to have a 2D array as a layout for
your data, where the row and column index would give the X and Y position
(scaled so that they fit in consecutive integers). With some luck you can
convert you V array to a properly-shaped 2D array simply using "reshape".
This correpsonds to the Z array in the code on the page you reference.

Hope this helps,

Ga?l


From pgmdevlist at gmail.com  Mon Nov  5 11:03:07 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Mon, 5 Nov 2007 11:03:07 -0500
Subject: [SciPy-user] time series analysis
In-Reply-To: <loom.20071105T131924-81@post.gmane.org>
References: <loom.20071105T131924-81@post.gmane.org>
Message-ID: <200711051103.07410.pgmdevlist@gmail.com>

Timmie,
> I want to use python to analyse and evaluate (measurement) time series.

The package TimeSeries was designed to perform the kind of operations you want 
(filtering, frequency conversion, handling missing data...). It is currently 
available in the sandbox of scipy. More info here:
http://www.scipy.org/SciPyPackages/TimeSeries

> * filling data gaps for days or hours without valid data using statistical
> methods like regression

This one might be a tad trickier. You may want to give pyloess a try, it's a 
wrapper around the dloess functions. Also available in the scipy sandbox.


From gael.varoquaux at normalesup.org  Mon Nov  5 13:45:13 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 5 Nov 2007 19:45:13 +0100
Subject: [SciPy-user] Kernel density approximation
Message-ID: <20071105184513.GK21399@clipper.ens.fr>

Hi,

What is the recommended way to compute a kernel density approximation of
a probability distribution function given a sample data ? My problem is
two dimensional, each data element is made of a pair of numbers.

I currently use an histogram2d, but I would like to use a smooth kernel,
such as a gaussian, and most important, I would like the algorithm to do
the work for me as far as choosing the bandwidth :->.

And I forgot to say that the function should be fast, as it is used in a
Monte Carlo algorithm, and is called many times.

Yes, I know, I want the Moon, on a stick, please :->.

Thanks,

Ga?l


From zpincus at stanford.edu  Mon Nov  5 13:58:36 2007
From: zpincus at stanford.edu (Zachary Pincus)
Date: Mon, 5 Nov 2007 13:58:36 -0500
Subject: [SciPy-user] Kernel density approximation
In-Reply-To: <20071105184513.GK21399@clipper.ens.fr>
References: <20071105184513.GK21399@clipper.ens.fr>
Message-ID: <F53E3CEB-A439-4332-BC71-BF478E80E7FD@stanford.edu>

This might be a start:
scipy.stats.kde.gaussian_kde

I use it regularly for my n-d KDE needs. I think it has all of the  
moon/stick features required.

Zach


On Nov 5, 2007, at 1:45 PM, Gael Varoquaux wrote:

> Hi,
>
> What is the recommended way to compute a kernel density  
> approximation of
> a probability distribution function given a sample data ? My  
> problem is
> two dimensional, each data element is made of a pair of numbers.
>
> I currently use an histogram2d, but I would like to use a smooth  
> kernel,
> such as a gaussian, and most important, I would like the algorithm  
> to do
> the work for me as far as choosing the bandwidth :->.
>
> And I forgot to say that the function should be fast, as it is used  
> in a
> Monte Carlo algorithm, and is called many times.
>
> Yes, I know, I want the Moon, on a stick, please :->.
>
> Thanks,
>
> Ga?l
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From gael.varoquaux at normalesup.org  Mon Nov  5 14:11:01 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 5 Nov 2007 20:11:01 +0100
Subject: [SciPy-user] Kernel density approximation
In-Reply-To: <F53E3CEB-A439-4332-BC71-BF478E80E7FD@stanford.edu>
References: <20071105184513.GK21399@clipper.ens.fr>
	<F53E3CEB-A439-4332-BC71-BF478E80E7FD@stanford.edu>
Message-ID: <20071105191101.GL21399@clipper.ens.fr>

On Mon, Nov 05, 2007 at 01:58:36PM -0500, Zachary Pincus wrote:
> This might be a start:
> scipy.stats.kde.gaussian_kde

wow ! Cool. I don't know how I managed to miss that, it was right under
my eyes, where I was looking, with a sensible name.

> I use it regularly for my n-d KDE needs. I think it has all of the  
> moon/stick features required.

Scipy really is a large arm useful for grabbing the moon and putting it
on a stick. Once again scipy rocks !

Thanks Zachary for the pointer, and thanks scipy devs for the function.

Ga?l


From timmichelsen at gmx-topmail.de  Mon Nov  5 14:24:24 2007
From: timmichelsen at gmx-topmail.de (Timmie)
Date: Mon, 5 Nov 2007 19:24:24 +0000 (UTC)
Subject: [SciPy-user] time series analysis
References: <loom.20071105T131924-81@post.gmane.org>
	<200711051103.07410.pgmdevlist@gmail.com>
Message-ID: <loom.20071105T192034-522@post.gmane.org>

Thanks for your answer. 

> The package TimeSeries was designed to perform the kind of operations you want 
> (filtering, frequency conversion, handling missing data...). It is currently 
> available in the sandbox of scipy. More info here:
> http://www.scipy.org/SciPyPackages/TimeSeries
Wow, this looks great. But a little complex ;-)
Well, one could write functions for common tasks that fascilitate it a bit...


> > * filling data gaps for days or hours without valid data using statistical
> > methods like regression
> 
> This one might be a tad trickier. You may want to give pyloess a try, it's a 
> wrapper around the dloess functions. Also available in the scipy sandbox.

Any idea when there will be a first binary release of all this together with
maskedarray?

Has anyone else done this kind of things?
Maybe with a interface to R?

Regards,
Timmie



From pgmdevlist at gmail.com  Mon Nov  5 14:37:33 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Mon, 5 Nov 2007 14:37:33 -0500
Subject: [SciPy-user] time series analysis
In-Reply-To: <loom.20071105T192034-522@post.gmane.org>
References: <loom.20071105T131924-81@post.gmane.org>
	<200711051103.07410.pgmdevlist@gmail.com>
	<loom.20071105T192034-522@post.gmane.org>
Message-ID: <200711051437.34362.pgmdevlist@gmail.com>

On Monday 05 November 2007 14:24:24 Timmie wrote:

> Wow, this looks great. But a little complex ;-)

> Well, one could write functions for common tasks that fascilitate it a
> bit...

You'd be surprised of how easy it is to use after a while. The conversion 
functions are quite useful. Please don't hesitate to contact em off-list if 
you have some special requests

> Any idea when there will be a first binary release of all this together
> with maskedarray?

Nope. The inclusion of maskedarray in numpy is still in the air, timeseries 
may become a scikit sooner or later.

> Has anyone else done this kind of things?
What kind of things ?

> Maybe with a interface to R?
I started that way (with the rpy package), but decided to stick to pure Python 
to avoid surprises...


From william.ratcliff at gmail.com  Mon Nov  5 15:28:55 2007
From: william.ratcliff at gmail.com (william ratcliff)
Date: Mon, 5 Nov 2007 15:28:55 -0500
Subject: [SciPy-user] weave and masked arrays?
Message-ID: <827183970711051228r77c2c871ne26c7abb893def23@mail.gmail.com>

Does anyone know if scipy.weave supports numpy masked arrays?

Thanks,
William


From s.mientki at ru.nl  Mon Nov  5 15:37:24 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Mon, 05 Nov 2007 21:37:24 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ? part 2
In-Reply-To: <4728FFC8.2070007@ru.nl>
References: <4728FFC8.2070007@ru.nl>
Message-ID: <472F7F04.1050905@ru.nl>

and here is the second part of the Vision demo:
  http://www.osc.edu/~unpingco/Tutorial_5Nov.html

I find this a very useful discussion,
not only for the good overview,
but also for what is possible now and will be possible in the future.

It also started me thinking:
"why do so many people (including me until now) find these graphical 
design packages,
like Vision / LabView so attractive / easy ? "

OK schematic designs in general read more easily than a piece of text,
for most people, watching TV is much more relaxing and attractive than 
reading a book.

But the main reason seems to me "a priori knowledge"
When I saw the second demo the "LabView" shivers came over me:
it's becoming too complex, there's too much behind the blocks that can't 
be seen.

Looking back at the first demo, I could also write that in one line of 
text (if I had the right libraries):

   Display_Image ( Rotate_Image ( Image_From_File ( 'demo.png' ) ) )

This line of code is easier to write then placing the blocks and 
connecting them in Vision.

Still I (and possible many others) find the Vision approach much more 
intuitive and much easier.
But there is another major difference than the graphics alone:
in the text version I need to know
- the exact name of each item: is it Rotate_Image, rotate_image, 
SetRotateImage, or ...
- the exact number of inputs and outputs
- the range and units of each input parameter: is rotation in 360 
degrees / 400 degrees or in radians ?

So I think we not necessarily  need a  graphical user interface,
but good libraries with easy to understand functions,
a priori given ranges and units, good help files etc.

So maybe not "graphical" is the magic word, but moreover "bricks"  like 
Lego uses them ;-)

Like to know what others think,
because I'm still very interested in getting non-programmers from 
LabView to SciPy,
and until now I couldn't convince any of the current LabView teachers / 
users :-(

cheers,
Stef




From jdh2358 at gmail.com  Mon Nov  5 16:09:34 2007
From: jdh2358 at gmail.com (John Hunter)
Date: Mon, 5 Nov 2007 15:09:34 -0600
Subject: [SciPy-user] Scipy + Vision = LabView ? part 2
In-Reply-To: <472F7F04.1050905@ru.nl>
References: <4728FFC8.2070007@ru.nl> <472F7F04.1050905@ru.nl>
Message-ID: <88e473830711051309v1556e3d5t579ef79a2f5e5e14@mail.gmail.com>

On Nov 5, 2007 2:37 PM, Stef Mientki <s.mientki at ru.nl> wrote:

>    Display_Image ( Rotate_Image ( Image_From_File ( 'demo.png' ) ) )

> So maybe not "graphical" is the magic word, but moreover "bricks"  like
> Lego uses them ;-)

I think for the most part we have the bricks: it's just that they are
scattered in many places w/o comprehensive documentation that
illustrates how to find them and put them together

  from scipy.misc import imread, imshow, imrotate
  imshow(imrotate(imread('moonlanding.jpg'), 20))

but how many know to look in scipy.misc?

JDH


From gael.varoquaux at normalesup.org  Mon Nov  5 17:16:43 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 5 Nov 2007 23:16:43 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ? part 2
In-Reply-To: <472F7F04.1050905@ru.nl>
References: <4728FFC8.2070007@ru.nl> <472F7F04.1050905@ru.nl>
Message-ID: <20071105221643.GB28991@clipper.ens.fr>

On Mon, Nov 05, 2007 at 09:37:24PM +0100, Stef Mientki wrote:
> So I think we not necessarily  need a  graphical user interface,
> but good libraries with easy to understand functions,
> a priori given ranges and units, good help files etc.

> So maybe not "graphical" is the magic word, but moreover "bricks"  like 
> Lego uses them ;-)

I think we need a good IDE, with integrated help/calltips, a good help, a
library browsers, a interactive debug for exploring program flow... I
wonder if SPE has a reusable API that we could use  to build this API,
along with envisage, ipython, once it has a wx front end, MPL,...

I must admit this is the dream plan that I would really like to see
happen in the next few years, and this is a lot what my pitch about
reusable components was about. If we have set of tools, all based on
WxWidget (just choosing WxWidget because it already has a lot of tools
available, and looks windowsish under windows), that we could assemble
for a good IDE (that means better than Matlab, it wouldn't be to hard to
do), something that would meld pycrust, ipython, MPL, mayavi (the tree on
the side is a good UI feature IMHO, that could be generalized), SPE...

I don't think the planets are yet aligned to start coding such a tools,
we are missing key components (e.g. wx front end to ipython), but if all
the people working on a tools think hard how they can make it a reusable
as possible, as easy to integrate seamlessly in an application, then
we'll get there in a year or two.


    Ga?l


From oliphant at enthought.com  Mon Nov  5 17:50:37 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Mon, 05 Nov 2007 16:50:37 -0600
Subject: [SciPy-user] weave and masked arrays?
In-Reply-To: <827183970711051228r77c2c871ne26c7abb893def23@mail.gmail.com>
References: <827183970711051228r77c2c871ne26c7abb893def23@mail.gmail.com>
Message-ID: <472F9E3D.3020605@enthought.com>

william ratcliff wrote:
> Does anyone know if scipy.weave supports numpy masked arrays?
>   
Not directly.  You should pass the data and mask arrays separately to a 
weave subroutine.

-Travis  O.




From mattknox_ca at hotmail.com  Mon Nov  5 22:47:41 2007
From: mattknox_ca at hotmail.com (Matt Knox)
Date: Tue, 6 Nov 2007 03:47:41 +0000 (UTC)
Subject: [SciPy-user] time series analysis
References: <loom.20071105T131924-81@post.gmane.org>
	<200711051103.07410.pgmdevlist@gmail.com>
	<loom.20071105T192034-522@post.gmane.org>
	<200711051437.34362.pgmdevlist@gmail.com>
Message-ID: <loom.20071106T033056-310@post.gmane.org>

> > Wow, this looks great. But a little complex 
> 
> > Well, one could write functions for common tasks that fascilitate it a
> > bit...

If you have any ideas for simplifying/improving things, we are certainly open 
to suggestions and would love the feedback. Being a sandbox package currently, 
there is no better time then now to get your ideas incorporated into the 
timeseries module.

> > Any idea when there will be a first binary release of all this together
> > with maskedarray?
> 
> Nope. The inclusion of maskedarray in numpy is still in the air, timeseries 
> may become a scikit sooner or later.

The timeseries package won't likely be moving anywhere until maskedarray moves 
somewhere else since it is dependent on it. I won't personally be providing 
any binaries while it is still in the sandbox either, but I am happy to 
provide advice on how to compile it on Windows.

Assuming maskedarray eventually moves into the core of numpy, my preference 
would be to put the timeseries module right into the main scipy trunk as I 
believe it is general purpose enough to warrant inclusion in scipy, and many 
people have expressed interest in it. But much discussion and debate will have 
to take place before that will happen.

Looking ahead really long term, I think built in support in matplotlib for 
TimeSeries objects (along the lines of the "plotlib" timeseries sub module) 
would be an improvement over the current approach to time series plotting in 
matplotlib too. But I have not discussed such things with any of the 
matplotlib developers and have no idea how they feel about that, and it is too 
early to discuss that yet anyway.

- Matt




From s.mientki at ru.nl  Tue Nov  6 03:26:37 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Tue, 06 Nov 2007 09:26:37 +0100
Subject: [SciPy-user] Scipy + Vision = LabView ?
In-Reply-To: <472CFBC4.1090802@ru.nl>
References: <4728FFC8.2070007@ru.nl>
	<40e64fa20710311536v68f351bbq4cbc7a836e7620bf@mail.gmail.com>
	<e76aa17f0710311551q53184dedva5e28e3e9d37b37b@mail.gmail.com>
	<20071102140821.GA1517@clipper.ens.fr> <472B6D19.50409@ru.nl>
	<20071103145233.GE1956@clipper.ens.fr> <472CFBC4.1090802@ru.nl>
Message-ID: <4730253D.5050105@ru.nl>


>>
>> So we have Vision, that is indeed impressive, but really built around
>> custom tools. It would be a large amount of work to abstract out the
>> visual programming part of the rest, but it is probably possible.
>>   
>>     
> Yes I took a quick look, and doesn't seems easy to extract the Vision 
> part form the total package.
>   
>
I asked a question about this in the Vision discussion list,
and according to the answer (given below) it shouldn't be too difficult:.
I tried this solution, but couldn't get it working,
because there seems something wrong with my tcl-package,
and now I remember, I have seen that before ;-)

cheers,
Stef

Vision can be easily extracted from the MGLTools package but it requires some additional packages.
MGLTools packages list critical and non-critical dependencies at the end of their __init__.py file.

Critical dependencies are the packages that are required for Vision to work at all.

Non critical dependencies will not prevent the package from working however, some functionality will
be missing.

The latest release of MGLTools (1.4.6) runs with python 2.4

It needs some 3rd party packages: Pmw, Numeric (23.8 or 24.2)

Inside MGLTools, the critical dependencies for Vision are:
Vision, NetworkEditor, mglutil, Support

an easy solution is to move thees 4 packages inside
/usr/lib/python2.4/site-packages/

then you can launch python
import Vision
Vision.runVision()

guillaume vareille



Het UMC St Radboud staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 41055629.
The Radboud University Nijmegen Medical Centre is listed in the Commercial Register of the Chamber of Commerce under file number 41055629.




From timmichelsen at gmx-topmail.de  Tue Nov  6 03:58:18 2007
From: timmichelsen at gmx-topmail.de (Timmie)
Date: Tue, 6 Nov 2007 08:58:18 +0000 (UTC)
Subject: [SciPy-user] time series analysis
References: <loom.20071105T131924-81@post.gmane.org>
	<200711051103.07410.pgmdevlist@gmail.com>
	<loom.20071105T192034-522@post.gmane.org>
	<200711051437.34362.pgmdevlist@gmail.com>
	<loom.20071106T033056-310@post.gmane.org>
Message-ID: <loom.20071106T083713-737@post.gmane.org>

> > > Wow, this looks great. But a little complex 
> > 
> > > Well, one could write functions for common tasks that fascilitate it a
> > > bit...
I believe you. Time is the matter.

> If you have any ideas for simplifying/improving things, we are certainly open 
> to suggestions and would love the feedback. Being a sandbox package currently, 
> there is no better time then now to get your ideas incorporated into the 
> timeseries module.
Well, since I am still a Python beginner somehow and still need to know how to
use it efficiently for my data analysis I will not really be able to contribute
to core.
At my current state of py knowledge I can alredy write my own simple modules
consisting of functions that I find useful.

Some things I can imagine are the following:
create a tools directory under the timeseries tree. then there we could place
things like
* common frequency conversions: reduce to hourly values
* error checking of measurement data: statistically and logically

> > > Any idea when there will be a first binary release of all this together
> > > with maskedarray?
> > 
> > Nope. The inclusion of maskedarray in numpy is still in the air, timeseries 
> > may become a scikit sooner or later.
Is it safe for me to replace/patch my current maskarray?

> The timeseries package won't likely be moving anywhere until maskedarray moves 
> somewhere else since it is dependent on it. I won't personally be providing 
> any binaries while it is still in the sandbox either, but I am happy to 
> provide advice on how to compile it on Windows.
I have experience in compiling on linux but have
to work on a windows box.
Therefore an advice on that would be useful.

Is there a possibility to subscibe to SVN to get an email on chnages?

Thanks for your responses.

Timmie



From mattknox_ca at hotmail.com  Tue Nov  6 06:38:17 2007
From: mattknox_ca at hotmail.com (Matt Knox)
Date: Tue, 6 Nov 2007 11:38:17 +0000 (UTC)
Subject: [SciPy-user] time series analysis
References: <loom.20071105T131924-81@post.gmane.org>
	<200711051103.07410.pgmdevlist@gmail.com>
	<loom.20071105T192034-522@post.gmane.org>
	<200711051437.34362.pgmdevlist@gmail.com>
	<loom.20071106T033056-310@post.gmane.org>
	<loom.20071106T083713-737@post.gmane.org>
Message-ID: <loom.20071106T111941-375@post.gmane.org>

> Some things I can imagine are the following:
> create a tools directory under the timeseries tree.

There is a "lib" sub-directory for stuff that falls outside the core 
Date/TimeSeries classes. It currently includes a sub-module for "moving 
functions" (moving average, etc...), and interpolation.

> * common frequency conversions: reduce to hourly values

Frequency conversions are simple to do using the "convert" method of the 
TimeSeries class. Here is an example converting an hourly frequency series to 
daily...

>>> import numpy as np
>>> import maskedarray as ma
>>> import timeseries as ts
>>> h = ts.time_series(np.arange(50, dtype=np.float32), start_date=ts.today
('hourly'))
>>> h
timeseries([  0.   1.   2.   3.   4.   5.   6.   7.   8.   9.  10.  11.  12.  
13.  14.
  15.  16.  17.  18.  19.  20.  21.  22.  23.  24.  25.  26.  27.  28.  29.
  30.  31.  32.  33.  34.  35.  36.  37.  38.  39.  40.  41.  42.  43.  44.
  45.  46.  47.  48.  49.],
           dates = [06-Nov-2007 06:00 ... 08-Nov-2007 07:00],
           freq  = H)

>>> d = h.convert('daily')
>>> d
timeseries(
 [[-- -- -- -- -- -- 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
  13.0 14.0 15.0 16.0 17.0]
 [18.0 19.0 20.0 21.0 22.0 23.0 24.0 25.0 26.0 27.0 28.0 29.0 30.0 31.0
  32.0 33.0 34.0 35.0 36.0 37.0 38.0 39.0 40.0 41.0]
 [42.0 43.0 44.0 45.0 46.0 47.0 48.0 49.0 -- -- -- -- -- -- -- -- -- -- --
  -- -- -- -- --]],
           dates =
 [06-Nov-2007 ... 08-Nov-2007],
           freq  = D)

>>> d_avg = h.convert('daily', ma.average)
>>> d_avg
timeseries([  8.5  29.5  45.5],
           dates = [06-Nov-2007 ... 08-Nov-2007],
           freq  = D)

===============================================
If any of the above seems mysterious, let me know and I can offer a more 
detailed explanation.

> * error checking of measurement data: statistically and logically

Some data error checking algorithms could be useful, yes. I won't likely be 
working on them in the near future though.

> Is it safe for me to replace/patch my current maskarray?

Generally speaking, the maskedarray package is mostly backwards compatible 
with the current numpy.ma package , and thus the api is very stable so you 
should be able to update it without any problems.

> I have experience in compiling on linux but have
> to work on a windows box.
> Therefore an advice on that would be useful.

If you are using Python 2.5 on windows, it is easy to do using mingw. If you 
are using an earlier version of Python, Visual studio 2003 is the easiest way 
to go.

> Is there a possibility to subscibe to SVN to get an email on chnages?

I don't know of any svn clients that have a built in way to do this, but it 
likely wouldn't be difficult to write your own script to do this since 
subversion has a command line interface.

- Matt



From elr1979 at gmail.com  Tue Nov  6 08:41:17 2007
From: elr1979 at gmail.com (Eduardo Rodrigues)
Date: Tue, 6 Nov 2007 10:41:17 -0300
Subject: [SciPy-user] Solve
References: <loom.20071105T131924-81@post.gmane.org><200711051103.07410.pgmdevlist@gmail.com><loom.20071105T192034-522@post.gmane.org><200711051437.34362.pgmdevlist@gmail.com><loom.20071106T033056-310@post.gmane.org><loom.20071106T083713-737@post.gmane.org>
	<loom.20071106T111941-375@post.gmane.org>
Message-ID: <000001c8207b$cebe66b0$a300a8c0@rodrigues>

Hi, I am starting in Python. I usually use Maple and I am with a problem to 
do the same project in Python.
I would like to solve an equation like x+y+1=0 for x. In Maple I use 
"solve(x+y+1=0,x);".
Do exist a similar command in Python/NumPy/SciPy?
Regards,
Eduardo. 



From mnandris at btinternet.com  Tue Nov  6 08:42:43 2007
From: mnandris at btinternet.com (Michael Nandris)
Date: Tue, 6 Nov 2007 13:42:43 +0000 (GMT)
Subject: [SciPy-user] Solve
In-Reply-To: <000001c8207b$cebe66b0$a300a8c0@rodrigues>
Message-ID: <965265.30328.qm@web86504.mail.ird.yahoo.com>

These examples for solving equations are in scipy_tutorial.pdf

from scipy import mat, linalg 

"""
Solves three simultaneous equations: 
    x + 3y + 5z = 10
    2x + 5y + z = 8
    2x + 3y + 8z =3
"""

XYparameters = mat('[1 3 5; 2 5 1; 2 3 8]')
constants = mat('[10; 8; 3]')
print linalg.solve( XYparameters, constants )
                                        # array([[-9.28],
                                        #        [ 5.16],
                                        #      [ 0.76]])

"""
Solves two simultaneous equations: 
    5x + 2y - 9 = 0         
    3x + 4y - 4 = 0 
"""

XYparameters = mat('[5 2; 3 4]')
constants = mat('[9; 4]')
print linalg.solve( XYparameters, constants )  # should print (2,-0.5)

Eduardo Rodrigues <elr1979 at gmail.com> wrote: Hi, I am starting in Python. I usually use Maple and I am with a problem to 
do the same project in Python.
I would like to solve an equation like x+y+1=0 for x. In Maple I use 
"solve(x+y+1=0,x);".
Do exist a similar command in Python/NumPy/SciPy?
Regards,
Eduardo. 

_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org
http://projects.scipy.org/mailman/listinfo/scipy-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071106/a7b0b2b4/attachment.html>

From Fabrice.Silva at crans.org  Tue Nov  6 08:53:54 2007
From: Fabrice.Silva at crans.org (Fabrice Silva)
Date: Tue, 06 Nov 2007 14:53:54 +0100
Subject: [SciPy-user] Solve
In-Reply-To: <965265.30328.qm@web86504.mail.ird.yahoo.com>
References: <000001c8207b$cebe66b0$a300a8c0@rodrigues>
	<965265.30328.qm@web86504.mail.ird.yahoo.com>
Message-ID: <1194357234.3135.26.camel@localhost>


Le mardi 06 novembre 2007 ? 13:42 +0000, Michael Nandris a ?crit :
> These examples for solving equations are in scipy_tutorial.pdf
> Solves three simultaneous equations: 
>     x + 3y + 5z = 10
>     2x + 5y + z = 8
>     2x + 3y + 8z =3
> """
> print linalg.solve( XYparameters, constants )


> Eduardo Rodrigues <elr1979 at gmail.com> wrote:
>         Hi, I am starting in Python. I usually use Maple and I am with
>         a problem to do the same project in Python.
>         I would like to solve an equation like x+y+1=0 for x. In Maple
>         I use "solve(x+y+1=0,x);".
>         Do exist a similar command in Python/NumPy/SciPy?

Isn't the request about symbolic computations, like in maple ?

-- 
Fabrice Silva <Fabrice.Silva at crans.org>




From gael.varoquaux at normalesup.org  Tue Nov  6 09:01:03 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Tue, 6 Nov 2007 15:01:03 +0100
Subject: [SciPy-user] Solve
In-Reply-To: <000001c8207b$cebe66b0$a300a8c0@rodrigues>
References: <loom.20071106T111941-375@post.gmane.org>
	<000001c8207b$cebe66b0$a300a8c0@rodrigues>
Message-ID: <20071106140103.GQ21399@clipper.ens.fr>

On Tue, Nov 06, 2007 at 10:41:17AM -0300, Eduardo Rodrigues wrote:
> Hi, I am starting in Python. I usually use Maple and I am with a problem to 
> do the same project in Python.
> I would like to solve an equation like x+y+1=0 for x. In Maple I use 
> "solve(x+y+1=0,x);".
> Do exist a similar command in Python/NumPy/SciPy?

Do you want a numericaly result, given the value of y, or do you want a
formal result ? The numerical value can be calculated using scipy, for
linear equation with the method exposed by Michael Nandris, for non
linear equations, you should look at scipy.optimize.zeros.

If you want a formal result, or a functional result, you can use sympy:

from sympy import solve, Symbol
x = Symbol('x')
y = Symbol('y')
f = solve(x+y+1, x) # f gives you the symbolic result
F = lambda v: f.subs(y, v)

F is a function that returns the exact expression of the root of you
equation. If you have several roots, than you need to do something like:

F = lambda v: [r.subs(y, v) for r in f]

Currently sympy is still yound, and won't be able to solve everythin you
throw at it.

HTH,

Ga?l


From elr1979 at gmail.com  Tue Nov  6 11:31:39 2007
From: elr1979 at gmail.com (Eduardo Rodrigues)
Date: Tue, 6 Nov 2007 13:31:39 -0300
Subject: [SciPy-user] Solve
References: <965265.30328.qm@web86504.mail.ird.yahoo.com>
Message-ID: <005501c82092$84736c10$a300a8c0@rodrigues>

If I want to solve an equation with linalg.solve I must know the equation, right? 

>From help:
"solve(a, b)
    Return the solution of a*x = b"

But if I don't know - or if I don't want to look - the equation for x, how can I solve it?
I will import equations from a file.

[]'s.




  ----- Original Message ----- 
  From: Michael Nandris 
  To: SciPy Users List 
  Sent: Tuesday, November 06, 2007 10:42 AM
  Subject: Re: [SciPy-user] Solve


  These examples for solving equations are in scipy_tutorial.pdf

  from scipy import mat, linalg 

  """
  Solves three simultaneous equations: 
      x + 3y + 5z = 10
      2x + 5y + z = 8
      2x + 3y + 8z =3
  """

  XYparameters = mat('[1 3 5; 2 5 1; 2 3 8]')
  constants = mat('[10; 8; 3]')
  print linalg.solve( XYparameters, constants )
                                          # array([[-9.28],
                                          #        [ 5.16],
                                          #      [ 0.76]])

  """
  Solves two simultaneous equations: 
      5x + 2y - 9 = 0         
      3x + 4y - 4 = 0 
  """

  XYparameters = mat('[5 2; 3 4]')
  constants = mat('[9; 4]')
  print linalg.solve( XYparameters, constants )  # should print (2,-0.5)

  Eduardo Rodrigues <elr1979 at gmail.com> wrote:
    Hi, I am starting in Python. I usually use Maple and I am with a problem to 
    do the same project in Python.
    I would like to solve an equation like x+y+1=0 for x. In Maple I use 
    "solve(x+y+1=0,x);".
    Do exist a similar command in Python/NumPy/SciPy?
    Regards,
    Eduardo. 

    _______________________________________________
    SciPy-user mailing list
    SciPy-user at scipy.org
    http://projects.scipy.org/mailman/listinfo/scipy-user





------------------------------------------------------------------------------


  _______________________________________________
  SciPy-user mailing list
  SciPy-user at scipy.org
  http://projects.scipy.org/mailman/listinfo/scipy-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071106/e18a2d87/attachment.html>

From lbolla at gmail.com  Tue Nov  6 10:45:39 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Tue, 6 Nov 2007 16:45:39 +0100
Subject: [SciPy-user] Solve
In-Reply-To: <005501c82092$84736c10$a300a8c0@rodrigues>
References: <965265.30328.qm@web86504.mail.ird.yahoo.com>
	<005501c82092$84736c10$a300a8c0@rodrigues>
Message-ID: <80c99e790711060745i47039a76k91b4d898083d1274@mail.gmail.com>

Maybe I'm misunderstanding, but it seems you look at solve(a,b) as a
function to return the solution of the linear symbolic equation: a * x = b
(whose solution is: x = b / a), with x scalar.
Instead, solve(a,b) gives the numerical solution of the linear system a * x
= b, with x and b 1d-arrays and a 2d-array. For examples, something like:
 2 * x1 + 3 * x2 = 4
4 * x1 + 5 * x2 = 6
where x = [x1, x2], b = [4, 6] and a = [[2, 3], [4, 5]]. solve(a,b) returns
[-1, 2].

If from the file you import a string with the symbolic representation of you
equation, and you want to solve it symbolically, stay with simpy.
hth,
L.

On 11/6/07, Eduardo Rodrigues <elr1979 at gmail.com> wrote:
>
>  If I want to solve an equation with linalg.solve I must know the
> equation, right?
>
> From help:
> "solve(a, b)
>     Return the solution of a*x = b"
>
>  But if I don't know - or if I don't want to look - the equation for x,
> how can I solve it?
> I will import equations from a file.
>
> []'s.
>
>
>
>
>
> ----- Original Message -----
> *From:* Michael Nandris <mnandris at btinternet.com>
> *To:* SciPy Users List <scipy-user at scipy.org>
> *Sent:* Tuesday, November 06, 2007 10:42 AM
> *Subject:* Re: [SciPy-user] Solve
>
>
> These examples for solving equations are in *scipy_tutorial.pdf*
>
> from scipy import mat, linalg
>
> """
> Solves three simultaneous equations:
>     x + 3y + 5z = 10
>     2x + 5y + z = 8
>     2x + 3y + 8z =3
> """
>
> XYparameters = mat('[1 3 5; 2 5 1; 2 3 8]')
> constants = mat('[10; 8; 3]')
> print linalg.solve( XYparameters, constants )
>                                         # array([[-9.28],
>                                         #        [ 5.16],
>                                         #      [ 0.76]])
>
> """
> Solves two simultaneous equations:
>     5x + 2y - 9 = 0
>     3x + 4y - 4 = 0
> """
>
> XYparameters = mat('[5 2; 3 4]')
> constants = mat('[9; 4]')
> print linalg.solve( XYparameters, constants )  # should print (2,-0.5)
>
> *Eduardo Rodrigues <elr1979 at gmail.com>* wrote:
>
> Hi, I am starting in Python. I usually use Maple and I am with a problem
> to
> do the same project in Python.
> I would like to solve an equation like x+y+1=0 for x. In Maple I use
> "solve(x+y+1=0,x);".
> Do exist a similar command in Python/NumPy/SciPy?
> Regards,
> Eduardo.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>  ------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071106/75b48a47/attachment.html>

From gael.varoquaux at normalesup.org  Tue Nov  6 11:01:56 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Tue, 6 Nov 2007 17:01:56 +0100
Subject: [SciPy-user] Solve
In-Reply-To: <005501c82092$84736c10$a300a8c0@rodrigues>
References: <965265.30328.qm@web86504.mail.ird.yahoo.com>
	<005501c82092$84736c10$a300a8c0@rodrigues>
Message-ID: <20071106160156.GV21399@clipper.ens.fr>

On Tue, Nov 06, 2007 at 01:31:39PM -0300, Eduardo Rodrigues wrote:
>    If I want to solve an equation with linalg.solve I must know the equation,
>    right?

>    From help:
>    "solve(a, b)
>        Return the solution of a*x = b"

>    But if I don't know - or if I don't want to look - the equation for x, how
>    can I solve it?
>    I will import equations from a file.

How are these equations defined ? Can you give us an example.

Cheers,

Ga?l


From pgmdevlist at gmail.com  Tue Nov  6 11:56:03 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 6 Nov 2007 11:56:03 -0500
Subject: [SciPy-user] Assessing the use of packages
Message-ID: <200711061156.03652.pgmdevlist@gmail.com>

All,

It's evaluation time in my department (Bio. & Ag. Engng, UGA), and I'd need to 
document the impact of my Python contributions on the scientific community at 
large, or more realistically on the numpy/scipy user community...

* Is there a way to estimate how many people installed one particular package 
from the SVN ?

* if it's too tricky, would anybody using maskedarray and/or timeseries and/or 
pyloess (the SVN packages I helped implementing) mind dropping me a line 
off-list, with a very short description of the main field of research/use ? 
I'd basically need some kind of numbers to give to the People-In-Charge. 

Thanks a lot in advance for your time.

P.

PS: Sorry for the cross-posting..


From marcbeck1982 at yahoo.com  Tue Nov  6 12:29:20 2007
From: marcbeck1982 at yahoo.com (Marc Beck)
Date: Tue, 6 Nov 2007 09:29:20 -0800 (PST)
Subject: [SciPy-user] gaussian smoothing
Message-ID: <919551.97062.qm@web52309.mail.re2.yahoo.com>

I am a graduate assistant at the Space Science Center at Morehead State University. I am currently writing a program to convert raw data from our 21m antenna into a *.FITS image using numpy and scipy. I need to convolve the pixels of my image. I have three arrays: one for the x- coordinate, one for the y- coordinate and one for the brightness of the pixel. I need to feed those three arrays into a function to convolve the pixels and I need to get the same three arrays out with the new values, or the rest of the program does not work. Someone from the python forum referred me to some code at http://www.scipy.org/Cookbook/SignalSmooth
   
  but I don't understand which variable is what and how to integrate this into my program. Please explain to me how I can make my program work. I hope that I have described clearly enough what I am planning to do.
   
  I e-mailed the author of that code and asked for help, but this person is very busy and does not have the time to help me.
   
  Marc Beck

 __________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071106/97e925e1/attachment.html>

From oliphant at enthought.com  Tue Nov  6 12:37:25 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Tue, 06 Nov 2007 11:37:25 -0600
Subject: [SciPy-user] gaussian smoothing
In-Reply-To: <919551.97062.qm@web52309.mail.re2.yahoo.com>
References: <919551.97062.qm@web52309.mail.re2.yahoo.com>
Message-ID: <4730A655.6070508@enthought.com>

Marc Beck wrote:
> I am a graduate assistant at the Space Science Center at Morehead 
> State University. I am currently writing a program to convert raw data 
> from our 21m antenna into a *.FITS image using numpy and scipy. I need 
> to convolve the pixels of my image. I have three arrays: one for the 
> x- coordinate, one for the y- coordinate and one for the brightness of 
> the pixel. I need to feed those three arrays into a function to 
> convolve the pixels and I need to get the same three arrays out with 
> the new values, or the rest of the program does not work. Someone from 
> the python forum referred me to some code at 
> http://www.scipy.org/Cookbook/SignalSmooth 
> <http://www.scipy.org/Cookbook/SignalSmooth>
>  
> but I don't understand which variable is what and how to integrate 
> this into my program. Please explain to me how I can make my program 
> work. I hope that I have described clearly enough what I am planning 
> to do.
>  
> I e-mailed the author of that code and asked for help, but this person 
> is very busy and does not have the time to help me.

Hi Marc,

We are going to need more information in order to help you.   
Convolution usually requires two inputs (the images or signals to be 
convolved) and it returns one output (the convolved signal).

What are the purposes of the x- and y- coordinates?   Is the data not 
evenly-sampled?    What do you expect to be done with these coordinates 
values?

You are not mentioning what you are convolving the input with?

To help you explore.  You can find general convolution as 
scipy.signal.convolve


Best regards,

--Travis O.





From as8ca at virginia.edu  Tue Nov  6 13:06:44 2007
From: as8ca at virginia.edu (Alok Singhal)
Date: Tue, 6 Nov 2007 13:06:44 -0500
Subject: [SciPy-user] gaussian smoothing
In-Reply-To: <919551.97062.qm@web52309.mail.re2.yahoo.com>
References: <919551.97062.qm@web52309.mail.re2.yahoo.com>
Message-ID: <20071106180643.GF21039@virginia.edu>

Hi,

On 06/11/07: 09:29, Marc Beck wrote:
> 
>    I am a graduate assistant at the Space Science Center at Morehead
>    State University. I am currently writing a program to convert raw data
>    from our 21m antenna into a *.FITS image using numpy and scipy. I need
>    to convolve the pixels of my image. I have three arrays: one for the
>    x- coordinate, one for the y- coordinate and one for the brightness of
>    the pixel. I need to feed those three arrays into a function to
>    convolve the pixels and I need to get the same three arrays out with
>    the new values, or the rest of the program does not work.

Here's one way to do it:

Assuming the brightness array is a two-dimensional array with shape
(nx, ny), where nx = len(x), ny = len(y); you can do something like:

# <brightness> is the two-dimensional array of brightness values,
# <sigma_x>, <sigma_y> are the sigma values for the two gaussians
# in x and y respectively.  NOTE: the sigma values are in terms of array
# indices, and not in the units of  x and y values.
#
# So, if x and y are uniformly spaced:
#
#    sigma_x = sigma_x / (x[1]-x[0])
#    sigma_y = sigma_y / (y[1]-y[0])
#
# The output image is in the array <out> in this example
from scipy.ndimage import gaussian_filter
gaussian_filter(brightness, sigma=(sigma_x, sigma_y), output=out, mode='constant', cval=0.0)

-Alok

-- 
Alok Singhal                               *   *          
Graduate Student, dept. of Astronomy   *           *     *
University of Virginia                                    
http://www.astro.virginia.edu/~as8ca/              *    * 


From timmichelsen at gmx-topmail.de  Tue Nov  6 14:36:25 2007
From: timmichelsen at gmx-topmail.de (Timmie)
Date: Tue, 6 Nov 2007 19:36:25 +0000 (UTC)
Subject: [SciPy-user] time series analysis
References: <loom.20071105T131924-81@post.gmane.org>
	<200711051103.07410.pgmdevlist@gmail.com>
	<loom.20071105T192034-522@post.gmane.org>
	<200711051437.34362.pgmdevlist@gmail.com>
	<loom.20071106T033056-310@post.gmane.org>
	<loom.20071106T083713-737@post.gmane.org>
	<loom.20071106T111941-375@post.gmane.org>
Message-ID: <loom.20071106T192712-36@post.gmane.org>

Matt Knox <mattknox_ca <at> hotmail.com> writes:

> 
> > Some things I can imagine are the following:
> > create a tools directory under the timeseries tree.
> 
> There is a "lib" sub-directory for stuff that falls outside the core 
> Date/TimeSeries classes. It currently includes a sub-module for "moving 
> functions" (moving average, etc...), and interpolation.
> 
> > * common frequency conversions: reduce to hourly values
> 
> Frequency conversions are simple to do using the "convert" method of the 
> TimeSeries class. Here is an example converting an hourly frequency series to 
> daily...
> 
> >>> import numpy as np
> >>> import maskedarray as ma
> >>> import timeseries as ts
> >>> h = ts.time_series(np.arange(50, dtype=np.float32), start_date=ts.today
> ('hourly'))
> >>> h
> timeseries([  0.   1.   2.   3.   4.   5.   6.   7.   8.   9.  10.  11.  12.  
> 13.  14.
>   15.  16.  17.  18.  19.  20.  21.  22.  23.  24.  25.  26.  27.  28.  29.
>   30.  31.  32.  33.  34.  35.  36.  37.  38.  39.  40.  41.  42.  43.  44.
>   45.  46.  47.  48.  49.],
>            dates = [06-Nov-2007 06:00 ... 08-Nov-2007 07:00],
>            freq  = H)
> 
> >>> d = h.convert('daily')
> >>> d
> timeseries(
>  [[-- -- -- -- -- -- 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
>   13.0 14.0 15.0 16.0 17.0]
>  [18.0 19.0 20.0 21.0 22.0 23.0 24.0 25.0 26.0 27.0 28.0 29.0 30.0 31.0
>   32.0 33.0 34.0 35.0 36.0 37.0 38.0 39.0 40.0 41.0]
>  [42.0 43.0 44.0 45.0 46.0 47.0 48.0 49.0 -- -- -- -- -- -- -- -- -- -- --
>   -- -- -- -- --]],
>            dates =
>  [06-Nov-2007 ... 08-Nov-2007],
>            freq  = D)
> 
> >>> d_avg = h.convert('daily', ma.average)
> >>> d_avg
> timeseries([  8.5  29.5  45.5],
>            dates = [06-Nov-2007 ... 08-Nov-2007],
>            freq  = D)
> 
> ===============================================
> If any of the above seems mysterious, let me know and I can offer a more 
> detailed explanation.
> 
> > * error checking of measurement data: statistically and logically
> 
> Some data error checking algorithms could be useful, yes. I won't likely be 
> working on them in the near future though.
I think that I will have a closer look at the package and then see how I can 
use it. Maybe I can contribute something or at least give you feedback.

Well, I am very happy that I found this package. I really uses some Google 
search but nothing helpful had turned out.

What I also fould (I think via the moin wiki at www.python.org):
* It is a Python package designed to accomplish some usual tasks during the 
analysis of climate variability using Python: http://www.pyclimate.org/
* CDAT makes use of an open-source, object-oriented, easy-to-learn scripting 
language (Python) to link together separate software subsystems and packages to 
form an integrated environment for data analysis. http://www-
pcmdi.llnl.gov/software-portal/cdat

These two packages seem to do quite similar tasks. Or at least head in the same 
direction. But they are depending mainly on the use of netCDF file storage. But 
maybe there could be some likeage?

Kind regards and thanks for your help,
Timmie

P.S.: You my also send me PM on this, too.



From pgmdevlist at gmail.com  Tue Nov  6 15:17:07 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 6 Nov 2007 15:17:07 -0500
Subject: [SciPy-user] time series analysis
In-Reply-To: <loom.20071106T192712-36@post.gmane.org>
References: <loom.20071105T131924-81@post.gmane.org>
	<loom.20071106T111941-375@post.gmane.org>
	<loom.20071106T192712-36@post.gmane.org>
Message-ID: <200711061517.07602.pgmdevlist@gmail.com>

Timmie,

Matt and I started TimeSeries each on our own, before merging our efforts last 
Christmas. On my side, I was trying to find something equivalent to pyclimate 
for my own purpose, but supporting numpy. I poked around this package a bit, 
wasn't completely happy with it, then figured it would be as easy to redo on 
top of numpy. It's while I was struggling with subclassing the masked arrays 
of numpy.core.ma that I decided to reimplement maskedarray, which eventually 
leads to the current version of timeseries.

If I'm not completely mistaken, numpy.core.ma is a translation via numeric of 
Paul Dubois' s implementation for CDAT. So yeah, I could have followed that 
path, but I was learning python at the same time and it was a very good 
exercise to reinvent the wheel, thus adding more noise and confusion.

I do agree that there could be some linkage between timeseries, pyclimate and 
CDAT, at least in terms of converting objects from package to another. 
However, it's unlikely to happen any time soon. Nevertheless, I may try to 
implement some of the functions of pyclimate and CDAT in numpy/scipy, when 
I'll have the need for them.

What I would suggest you is to start with timeseries, as it's pretty easy to 
use. Then, depending on what your needs are, you can start developing your 
own functions.

In any case, don't hesitate to contact Matt or myself if you need some 
specific help. The TimeSeriesPackage page of the scipy wiki is a good start, 
Matt did a terrific job.

Sincerely
P.


From brian.clowers at pnl.gov  Tue Nov  6 17:53:20 2007
From: brian.clowers at pnl.gov (Clowers, Brian H)
Date: Tue, 6 Nov 2007 14:53:20 -0800
Subject: [SciPy-user] HDF5 Viewer for Python
In-Reply-To: <mailman.6407.1194363324.2308.scipy-user@scipy.org>
References: <mailman.6407.1194363324.2308.scipy-user@scipy.org>
Message-ID: <C4BF0852F93C9E40AFD9338575B89658015C8DC3@EMAIL01.pnl.gov>

I have data sets in a HDF5 format that contain a number of arrays which
I'd like to plot using matplotlib.  I can do this pretty easily using
just IDLE and pyTables, however, it would be nice to put together a GUI
and automate some of the key strokes.  Does anyone know of or have an
alternative to viTables that they'd be willing to share?  viTables looks
great but it is also 250 Euros for single license--then again if it had
the capability to directly plot to matplotlib I'd consider it.  Anyway,
when it comes to a GUI I imagine the challenge would be how to link the
arrays or tables in the HDF5 file to a Treeview in an interactive
manner.  Any thoughts?  

Thanks,

Brian


From R.Springuel at umit.maine.edu  Tue Nov  6 18:49:56 2007
From: R.Springuel at umit.maine.edu (R. Padraic Springuel)
Date: Tue, 06 Nov 2007 18:49:56 -0500
Subject: [SciPy-user] Element ranks for an array
Message-ID: <4730FDA4.1070107@umit.maine.edu>

Does scipy have a function that can provide the element ranks by size 
for a 1-d array?

I.e. a function that would return [3,4,2,1] or [2,1,3,4] (depending on 
ascending or descending ranking) for array([50, 10, 62, 1000]).
-- 

R. Padraic Springuel
Teaching Assistant
Department of Physics and Astronomy
University of Maine
Bennett 309
Office Hours: By appointment only


From pgmdevlist at gmail.com  Tue Nov  6 19:10:41 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 6 Nov 2007 19:10:41 -0500
Subject: [SciPy-user] Element ranks for an array
In-Reply-To: <4730FDA4.1070107@umit.maine.edu>
References: <4730FDA4.1070107@umit.maine.edu>
Message-ID: <200711061910.42609.pgmdevlist@gmail.com>

On Tuesday 06 November 2007 18:49:56 R. Padraic Springuel wrote:
> Does scipy have a function that can provide the element ranks by size
> for a 1-d array?
>
> I.e. a function that would return [3,4,2,1] or [2,1,3,4] (depending on
> ascending or descending ranking) for array([50, 10, 62, 1000]).

For a 1D array
numpy.arange(1,len(x)+1)[x.argsort()]

or simply
scipy.stats.rankdata(x)

That'll give you ranking by descending order. Use the [::-1] to get the 
reverse.


From as8ca at virginia.edu  Tue Nov  6 19:20:37 2007
From: as8ca at virginia.edu (Alok Singhal)
Date: Tue, 6 Nov 2007 19:20:37 -0500
Subject: [SciPy-user] Element ranks for an array
In-Reply-To: <4730FDA4.1070107@umit.maine.edu>
References: <4730FDA4.1070107@umit.maine.edu>
Message-ID: <20071107002036.GA23267@virginia.edu>

Hi,

On 06/11/07: 18:49, R. Padraic Springuel wrote:
> Does scipy have a function that can provide the element ranks by size 
> for a 1-d array?
> 
> I.e. a function that would return [3,4,2,1] or [2,1,3,4] (depending on 
> ascending or descending ranking) for array([50, 10, 62, 1000]).

I assume you meant [4, 3, 1, 2] in the first case.

Look at argsort() in numpy.

This should do what you want:

def rank(x, ascending=True):
	i = numpy.argsort(x) + 1
	if ascending:
		return i
	else:
		return i[::-1]

-Alok

-- 
Alok Singhal                               *   *          
Graduate Student, dept. of Astronomy   *           *     *
University of Virginia                                    
http://www.astro.virginia.edu/~as8ca/              *    * 


From robert.kern at gmail.com  Tue Nov  6 19:30:24 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 06 Nov 2007 18:30:24 -0600
Subject: [SciPy-user] Element ranks for an array
In-Reply-To: <20071107002036.GA23267@virginia.edu>
References: <4730FDA4.1070107@umit.maine.edu>
	<20071107002036.GA23267@virginia.edu>
Message-ID: <47310720.3020305@gmail.com>

Alok Singhal wrote:
> Hi,
> 
> On 06/11/07: 18:49, R. Padraic Springuel wrote:
>> Does scipy have a function that can provide the element ranks by size 
>> for a 1-d array?
>>
>> I.e. a function that would return [3,4,2,1] or [2,1,3,4] (depending on 
>> ascending or descending ranking) for array([50, 10, 62, 1000]).
> 
> I assume you meant [4, 3, 1, 2] in the first case.
> 
> Look at argsort() in numpy.
> 
> This should do what you want:
> 
> def rank(x, ascending=True):
> 	i = numpy.argsort(x) + 1
> 	if ascending:
> 		return i
> 	else:
> 		return i[::-1]

No, this is incorrect. He correctly wanted [3,4,2,1] for the descending ranking.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From as8ca at virginia.edu  Tue Nov  6 20:12:05 2007
From: as8ca at virginia.edu (Alok Singhal)
Date: Tue, 6 Nov 2007 20:12:05 -0500
Subject: [SciPy-user] Element ranks for an array
In-Reply-To: <47310720.3020305@gmail.com>
References: <4730FDA4.1070107@umit.maine.edu>
	<20071107002036.GA23267@virginia.edu> <47310720.3020305@gmail.com>
Message-ID: <20071107011205.GA23486@virginia.edu>

On 06/11/07: 18:30, Robert Kern wrote:
> No, this is incorrect. He correctly wanted [3,4,2,1] for the
> descending ranking.

Oops.  I misunderstood the question and gave the wrong answer.

Sorry,
Alok

-- 
Alok Singhal                               *   *          
Graduate Student, dept. of Astronomy   *           *     *
University of Virginia                                    
http://www.astro.virginia.edu/~as8ca/              *    * 


From gael.varoquaux at normalesup.org  Wed Nov  7 02:37:50 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Wed, 7 Nov 2007 08:37:50 +0100
Subject: [SciPy-user] HDF5 Viewer for Python
In-Reply-To: <C4BF0852F93C9E40AFD9338575B89658015C8DC3@EMAIL01.pnl.gov>
References: <mailman.6407.1194363324.2308.scipy-user@scipy.org>
	<C4BF0852F93C9E40AFD9338575B89658015C8DC3@EMAIL01.pnl.gov>
Message-ID: <20071107073750.GD27897@clipper.ens.fr>

On Tue, Nov 06, 2007 at 02:53:20PM -0800, Clowers, Brian H wrote:
> I have data sets in a HDF5 format that contain a number of arrays which
> I'd like to plot using matplotlib.  I can do this pretty easily using
> just IDLE and pyTables, however, it would be nice to put together a GUI
> and automate some of the key strokes.  Does anyone know of or have an
> alternative to viTables that they'd be willing to share?  viTables looks
> great but it is also 250 Euros for single license--then again if it had
> the capability to directly plot to matplotlib I'd consider it.  Anyway,
> when it comes to a GUI I imagine the challenge would be how to link the
> arrays or tables in the HDF5 file to a Treeview in an interactive
> manner.  Any thoughts?  

I wacked something specific for the hdf5 structure of the data of
my experiment together with pytables, matplotlib, and traits in 2 day's
work. It's ugly (really) and terribly specific to my data structure, but
it shows that doing something quick and dirty isn't hard.

I am not to sure I am willing to share the code, (:->, I will, if
somebody _really_ asks for it) because I think I could just puke by
looking at it. It came from an immediate need, not thought at all in
advance, with deadlines right at the corner, and no intention of reuse,
so it is horrendous. 

Something pretty would be possible with a bit more work, but not that
much, and by putting some actual thinking in it, rather than just coding
something strait up. I would use the same tools: pytables, matplotlib and
traits.

Ga?l


From emin.shopper at gmail.com  Wed Nov  7 10:10:04 2007
From: emin.shopper at gmail.com (Emin.shopper Martinian.shopper)
Date: Wed, 7 Nov 2007 10:10:04 -0500
Subject: [SciPy-user] Is this a bug: scipy.cov([0,1]) = 0
Message-ID: <32e43bb70711070710x65c42568rbdc4942743af534b@mail.gmail.com>

Dear Experts,

I recently noticed some strange behavior with scipy (and numpy as
well). Doing scipy.cov([0,1]) returns 0 which I believe is incorrect.
Doing scipy.cov([0.0,1.0]) returns a more reasonable answer so I
suspect that the problem is that scipy is using integer math.

Is this considered a bug?

I can see no good reason to do integer math when computing a covariance.

Thanks,
-Emin


From elcorto at gmx.net  Wed Nov  7 10:36:33 2007
From: elcorto at gmx.net (Steve Schmerler)
Date: Wed, 07 Nov 2007 16:36:33 +0100
Subject: [SciPy-user] Is this a bug: scipy.cov([0,1]) = 0
In-Reply-To: <32e43bb70711070710x65c42568rbdc4942743af534b@mail.gmail.com>
References: <32e43bb70711070710x65c42568rbdc4942743af534b@mail.gmail.com>
Message-ID: <4731DB81.9050800@gmx.net>

Emin.shopper Martinian.shopper wrote:
> Dear Experts,
> 
> I recently noticed some strange behavior with scipy (and numpy as
> well). Doing scipy.cov([0,1]) returns 0 which I believe is incorrect.
> Doing scipy.cov([0.0,1.0]) returns a more reasonable answer so I
> suspect that the problem is that scipy is using integer math.
> 
> Is this considered a bug?
> 
> I can see no good reason to do integer math when computing a covariance.
> 

Works here.

In [14]: scipy.stats.cov([0,1])
Out[14]: 0.5

In [15]: scipy.cov([0,1])
Out[15]: array(0.5)

In [16]: numpy.cov([0,1])
Out[16]: array(0.5)

In [17]: numpy.cov([0.,1.])
Out[17]: array(0.5)

In [18]: scipy.cov([0.,1.])
Out[18]: array(0.5)

In [19]: scipy.stats.cov([0.,1.])
Out[19]: 0.5

In [20]: scipy.__version__
Out[20]: '0.7.0.dev3498'

In [21]: numpy.__version__
Out[21]: '1.0.5.dev4407'


-- 
cheers,
steve

Random number generation is the art of producing pure gibberish as quickly as 
possible.


From j.reid at mail.cryst.bbk.ac.uk  Wed Nov  7 11:03:06 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Wed, 07 Nov 2007 16:03:06 +0000
Subject: [SciPy-user] Undefined variable bug in
	scipy.stats.gaussian_kde.integrate_1d
Message-ID: <fgsnl3$nn9$2@ger.gmane.org>

I get the following error when calling scipy.stats.gaussian_kde.integrate_1d

c:\apps\python25\lib\site-packages\scipy\stats\kde.py in 
integrate_box_1d(self, low, high)
     204             raise ValueError("integrate_box_1d() only handles 
1D pdfs")
     205
--> 206         stdev = np.ravel(sqrt(self.covariance))[0]
     207
     208         normalized_low = ravel((low - self.dataset)/stdev)

<type 'exceptions.NameError'>: global name 'np' is not defined
 > 
c:\apps\python25\lib\site-packages\scipy\stats\kde.py(206)integrate_box_1d()
     205
--> 206         stdev = np.ravel(sqrt(self.covariance))[0]
     207

I had a quick look at the source and I think perhaps np should be numpy. 
Also ravel is explicitly imported from numpy in the file.

John.



From brian.clowers at pnl.gov  Wed Nov  7 12:49:34 2007
From: brian.clowers at pnl.gov (Clowers, Brian H)
Date: Wed, 7 Nov 2007 09:49:34 -0800
Subject: [SciPy-user] HDF5 Viewer for Python
Message-ID: <C4BF0852F93C9E40AFD9338575B89658015C8DC5@EMAIL01.pnl.gov>

Gael,

Thanks for the input.  A while back I dabbled with Traits and look at
your tutorial--thanks by the way.  However, I was unsure how to use
traits to develop some sort of Treeview for the HDF5 format.  Which
leads me to beg for your help...I'm sure your 'ugly' code is better than
mine on just about any given day.  If you'd be willing to share--even if
it is to get me started I'd appreciate your help.  The enthought packed
looks very promising so now might be a good time for me to take it up.
I upgraded my system to use the latest version of traits (I believe it
is 3) but am unsure how to begin populating a tree.  Any examples would
be great.  

Thanks

Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071107/2a7d69a9/attachment.html>

From robert.kern at gmail.com  Wed Nov  7 13:04:05 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 07 Nov 2007 12:04:05 -0600
Subject: [SciPy-user] Undefined variable bug
	in	scipy.stats.gaussian_kde.integrate_1d
In-Reply-To: <fgsnl3$nn9$2@ger.gmane.org>
References: <fgsnl3$nn9$2@ger.gmane.org>
Message-ID: <4731FE15.7050503@gmail.com>

John Reid wrote:
> I get the following error when calling scipy.stats.gaussian_kde.integrate_1d
> 
> c:\apps\python25\lib\site-packages\scipy\stats\kde.py in 
> integrate_box_1d(self, low, high)
>      204             raise ValueError("integrate_box_1d() only handles 
> 1D pdfs")
>      205
> --> 206         stdev = np.ravel(sqrt(self.covariance))[0]
>      207
>      208         normalized_low = ravel((low - self.dataset)/stdev)
> 
> <type 'exceptions.NameError'>: global name 'np' is not defined
>  > 
> c:\apps\python25\lib\site-packages\scipy\stats\kde.py(206)integrate_box_1d()
>      205
> --> 206         stdev = np.ravel(sqrt(self.covariance))[0]
>      207
> 
> I had a quick look at the source and I think perhaps np should be numpy. 
> Also ravel is explicitly imported from numpy in the file.

When reporting errors, please state the version which you are using. In this
case, the errors have already been fixed in SVN.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From s.mientki at ru.nl  Wed Nov  7 13:14:54 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Wed, 07 Nov 2007 19:14:54 +0100
Subject: [SciPy-user] Is Elefant alive ?
Message-ID: <4732009E.1010605@ru.nl>

hello,

Does anyone knows how I can reach one of the Elefant designers:
  http://elefant.developer.nicta.com.au/

I've spitted through all their mailinglists ;-)
but couldn't find an address.

I also subscribed to their mailing lists,
but didn't hear anything for 2 days ;-)

thanks,
Stef Mientki


From matthieu.brucher at gmail.com  Wed Nov  7 13:20:24 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Wed, 7 Nov 2007 19:20:24 +0100
Subject: [SciPy-user] Is Elefant alive ?
In-Reply-To: <4732009E.1010605@ru.nl>
References: <4732009E.1010605@ru.nl>
Message-ID: <e76aa17f0711071020v71a5bbf5of21d2da7676060b3@mail.gmail.com>

Did you try : admin at lineal.developer.nicta.com.au ?

Matthieu

2007/11/7, Stef Mientki <s.mientki at ru.nl>:
>
> hello,
>
> Does anyone knows how I can reach one of the Elefant designers:
>   http://elefant.developer.nicta.com.au/
>
> I've spitted through all their mailinglists ;-)
> but couldn't find an address.
>
> I also subscribed to their mailing lists,
> but didn't hear anything for 2 days ;-)
>
> thanks,
> Stef Mientki
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071107/743b1272/attachment.html>

From stef.mientki at gmail.com  Wed Nov  7 13:32:29 2007
From: stef.mientki at gmail.com (stef mientki)
Date: Wed, 07 Nov 2007 19:32:29 +0100
Subject: [SciPy-user] Is Elefant alive ?
In-Reply-To: <e76aa17f0711071020v71a5bbf5of21d2da7676060b3@mail.gmail.com>
References: <4732009E.1010605@ru.nl>
	<e76aa17f0711071020v71a5bbf5of21d2da7676060b3@mail.gmail.com>
Message-ID: <473204BD.2000203@gmail.com>

Matthieu Brucher wrote:
> Did you try : admin at lineal.developer.nicta.com.au 
> <http://lineal.developer.nicta.com.au> ?
thanks Matthieu,
their email addresses seems to be scares, but I found one,
now wait again ;-)

cheers,
Stef



From markbak at gmail.com  Wed Nov  7 15:37:14 2007
From: markbak at gmail.com (Mark Bakker)
Date: Wed, 7 Nov 2007 21:37:14 +0100
Subject: [SciPy-user] Geostatistics with Scipy?
Message-ID: <6946b9500711071237k17d76122ta2dc9e7fe505c0d8@mail.gmail.com>

Hello -

I have been looking for some package that can doe Geostatistics with Scipy.
Mostly Kriging of 2D data.

I expect somebody has done something like this before.

Does anybody know of some code somewhere?

Thanks,

Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071107/8d908b02/attachment.html>

From emin.shopper at gmail.com  Wed Nov  7 15:37:28 2007
From: emin.shopper at gmail.com (Emin.shopper Martinian.shopper)
Date: Wed, 7 Nov 2007 15:37:28 -0500
Subject: [SciPy-user] Is this a bug: scipy.cov([0,1]) = 0
In-Reply-To: <4731DB81.9050800@gmx.net>
References: <32e43bb70711070710x65c42568rbdc4942743af534b@mail.gmail.com>
	<4731DB81.9050800@gmx.net>
Message-ID: <32e43bb70711071237u78c7656bie3aef20d774c99cb@mail.gmail.com>

Ahh, good point. I was using numpy version 1.0 and scipy version
0.5.2. I guess its time to upgrade.

Thanks,
-Emin

On Nov 7, 2007 10:36 AM, Steve Schmerler <elcorto at gmx.net> wrote:
>
> Emin.shopper Martinian.shopper wrote:
> > Dear Experts,
> >
> > I recently noticed some strange behavior with scipy (and numpy as
> > well). Doing scipy.cov([0,1]) returns 0 which I believe is incorrect.
> > Doing scipy.cov([0.0,1.0]) returns a more reasonable answer so I
> > suspect that the problem is that scipy is using integer math.
> >
> > Is this considered a bug?
> >
> > I can see no good reason to do integer math when computing a covariance.
> >
>
> Works here.
>
> In [14]: scipy.stats.cov([0,1])
> Out[14]: 0.5
>
> In [15]: scipy.cov([0,1])
> Out[15]: array(0.5)
>
> In [16]: numpy.cov([0,1])
> Out[16]: array(0.5)
>
> In [17]: numpy.cov([0.,1.])
> Out[17]: array(0.5)
>
> In [18]: scipy.cov([0.,1.])
> Out[18]: array(0.5)
>
> In [19]: scipy.stats.cov([0.,1.])
> Out[19]: 0.5
>
> In [20]: scipy.__version__
> Out[20]: '0.7.0.dev3498'
>
> In [21]: numpy.__version__
> Out[21]: '1.0.5.dev4407'
>
>
> --
> cheers,
> steve
>
> Random number generation is the art of producing pure gibberish as quickly as
> possible.
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From william.ratcliff at gmail.com  Wed Nov  7 17:55:32 2007
From: william.ratcliff at gmail.com (william ratcliff)
Date: Wed, 7 Nov 2007 17:55:32 -0500
Subject: [SciPy-user] openmp and weave?
Message-ID: <827183970711071455m29f4869bo8c8312e0eca79a6e@mail.gmail.com>

Has anyone had any luck using weave with openMP?  If so, what did you
have to do?  I've started by updating my compiler to MinGW:
gcc-4.2.1-dw-2-2 (and similarly for g++), but am running into problems
with code written in weave that doesn't use any of openmp:

Here is the code:

import numpy as N
import weave
from weave import converters



def blitz_interpolate(x,y):


    code = """
    int pts = Ny[0];
//#pragma omp parallel for
    for (int i=0; i < pts-1; i++){
        y(i) = sin( exp( cos( - exp( sin(x(i)) ) ) ) );
     }
    return_val = 3;
    """
    extra=["-fopenmp -Lc:/python25/ -lPthreadGC2"]
    extra=[]
    z=weave.inline(code,['x','y'],type_converters=converters.blitz,compiler='gcc')
    print z
    return





if __name__=="__main__":
    x=N.arange(1000)
    y=N.zeros(x.shape,'d')
    blitz_interpolate(x,y)
    print x[35], y[35],N.sin(N.exp(N.cos(-N.exp(N.sin(x[35])))))





This works fine with version 3.4.2 of gcc, g++

Thanks,
William


From lev at vpac.org  Wed Nov  7 18:08:08 2007
From: lev at vpac.org (Lev Lafayette)
Date: Thu, 08 Nov 2007 10:08:08 +1100
Subject: [SciPy-user] SciPy test fails..
Message-ID: <1194476888.22909.1.camel@sys09.in.vpac.org>


Hey people,

I've just installed NUMPY and SciPy on our cluster and the standard test
fails. Can someone give me some pointers on what's wrong?

The relevant sections of the test follows.

Thanks in advance,



Lev



  Found 128 tests for scipy.lib.blas.fblas

****************************************************************
WARNING: clapack module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses flapack instead of clapack.
****************************************************************

  Found 7 tests for scipy.linalg.matfuncs
Warning: FAILURE importing tests for <module 'scipy.linsolve.umfpack'
from '.../linsolve/umfpack/__init__.pyc'>
/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/scipy/linsolve/umfpack/tests/test_umfpack.py:17: AttributeError: 'module' object has no attribute 'umfpack' (in ?)
Warning: FAILURE importing tests for <module
'scipy.linsolve.umfpack.umfpack' from
'...y/linsolve/umfpack/umfpack.pyc'>
/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/scipy/linsolve/umfpack/tests/test_umfpack.py:17: AttributeError: 'module' object has no attribute 'umfpack' (in ?)
  Found 0 tests for __main__
.../usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/scipy/cluster/vq.py:477: UserWarning: One of the clusters is empty. Re-run kmean with a different initialization.
  warnings.warn("One of the clusters is empty. "
exception raised as expected: One of the clusters is empty. Re-run kmean
with a different initialization.
................................................Residual:
1.05006987327e-07
..................../usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/scipy/interpolate/fitpack2.py:458: UserWarning: 
The coefficients of the spline returned have been computed as the
minimal norm least-squares solution of a (numerically) rank deficient
system (deficiency=7). If deficiency is large, the results may be
inaccurate. Deficiency may strongly depend on the value of eps.
  warnings.warn(message)
......
Don't worry about a warning regarding the number of bytes read.
..zswap:n=3
.....................................................................................
****************************************************************
WARNING: cblas module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses fblas instead of cblas.
****************************************************************

...........................................................................................caxpy:n=4
..zswap:n=3
..........
****************************************************************
WARNING: clapack module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses flapack instead of clapack.
****************************************************************

...Result may be inaccurate, approximate err = 1.73156859734e-08
...Result may be inaccurate, approximate err = 7.73070496507e-12
............................................................................................................/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/scipy/ndimage/interpolation.py:41: UserWarning: Mode "reflect" may yield incorrect results on boundaries. Please use "mirror" instead.
  warnings.warn('Mode "reflect" may yield incorrect results on '
........................................................................................................................................................................................................................................................................................................F..F.........................................................Use minimum degree ordering on A'+A.
.....................................Use minimum degree ordering on
A'+A.
.....................................Use minimum degree ordering on
A'+A.
................................Use minimum degree ordering on A'+A.
....................................................................................................................................................................................................................................................................................................................................................0.2
0.2
0.2
......0.2
..0.2
0.2
0.2
0.2
0.2
.........................................................................................................................................................................................................Ties preclude use of exact statistic.
..Ties preclude use of exact statistic.
......
======================================================================
FAIL: test_explicit (scipy.tests.test_odr.test_odr)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/scipy/odr/tests/test_odr.py", line 49, in test_explicit
    np.array([  1.2646548050648876e+03,  -5.4018409956678255e+01,
  File
"/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/numpy/testing/utils.py", line 230, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File
"/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/numpy/testing/utils.py", line 215, in assert_array_compare
    assert cond, msg
AssertionError: 
Arrays are not almost equal

(mismatch 100.0%)
 x: array([  1.26462971e+03,  -5.42545890e+01,  -8.64250389e-02])
 y: array([  1.26465481e+03,  -5.40184100e+01,  -8.78497122e-02])

======================================================================
FAIL: test_multi (scipy.tests.test_odr.test_odr)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/scipy/odr/tests/test_odr.py", line 190, in test_multi
    np.array([ 4.3799880305938963,  2.4333057577497703,
8.0028845899503978,
  File
"/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/numpy/testing/utils.py", line 230, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File
"/usr/local/python/2.4.4-gcc/lib/python2.4/site-packages/numpy/testing/utils.py", line 215, in assert_array_compare
    assert cond, msg
AssertionError: 
Arrays are not almost equal

(mismatch 100.0%)
 x: array([ 4.31272063,  2.44289312,  7.76215871,  0.55995622,
0.46423343])
 y: array([ 4.37998803,  2.43330576,  8.00288459,  0.51011472,
0.51739023])

----------------------------------------------------------------------
Ran 1716 tests in 3.670s

FAILED (failures=2)



From robert.kern at gmail.com  Wed Nov  7 18:09:45 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 07 Nov 2007 17:09:45 -0600
Subject: [SciPy-user] SciPy test fails..
In-Reply-To: <1194476888.22909.1.camel@sys09.in.vpac.org>
References: <1194476888.22909.1.camel@sys09.in.vpac.org>
Message-ID: <473245B9.2040202@gmail.com>

Lev Lafayette wrote:
> Hey people,
> 
> I've just installed NUMPY and SciPy on our cluster and the standard test
> fails. Can someone give me some pointers on what's wrong?

These are fixed in SVN.

http://scipy.org/scipy/scipy/ticket/357

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From lev at vpac.org  Wed Nov  7 18:31:56 2007
From: lev at vpac.org (Lev Lafayette)
Date: Thu, 08 Nov 2007 10:31:56 +1100
Subject: [SciPy-user] SciPy test fails..
In-Reply-To: <473245B9.2040202@gmail.com>
References: <1194476888.22909.1.camel@sys09.in.vpac.org>
	<473245B9.2040202@gmail.com>
Message-ID: <1194478316.22909.3.camel@sys09.in.vpac.org>

On Wed, 2007-11-07 at 17:09 -0600, Robert Kern wrote:
> Lev Lafayette wrote:
> > Hey people,
> > 
> > I've just installed NUMPY and SciPy on our cluster and the standard test
> > fails. Can someone give me some pointers on what's wrong?
> 
> These are fixed in SVN.
> 
> http://scipy.org/scipy/scipy/ticket/357
> 

Ahh, brilliant! Thank you!



From samrobertsmith at gmail.com  Wed Nov  7 19:49:13 2007
From: samrobertsmith at gmail.com (linda.s)
Date: Wed, 7 Nov 2007 17:49:13 -0700
Subject: [SciPy-user] about MLab
In-Reply-To: <472DF7C9.8020305@enthought.com>
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com>
	<472DF7C9.8020305@enthought.com>
Message-ID: <1d987df30711071649j4fd428j4aa999a04c6aea2a@mail.gmail.com>

On 11/4/07, Travis E. Oliphant <oliphant at enthought.com> wrote:
> linda.s wrote:
> > I have quite a few codes written in Numeric.
> > MLab comes with Numeric too.
> > I got confused when I began to use numpy because it does not have MLab
> > and does not have some same functions in Numeric.
> > is there any smooth way to change the code developed in Numeric to numpy?
> >
>
> You should also read the first chapters of the book "Guide to NumPy"
> which are available on-line for immediate download.  There is a
> comprehensive description of the differences between Numeric and NumPy
> as well as tips for compatibility.
>
> You must also import numpy.oldnumeric before it is available.
>
> import numpy.oldnumeric
>
> numpy.oldnumeric.mlab
>
> You will notice that the module is simple a re-naming module.  All of
> the functionality is in numpy but under possibly different names.
>
> All functions in Numeric are available in NumPy (as far as I know).  If
> you encounter a situation where that is not true, please let me know.
>
> -Travis

I still have the problem such as:

>>> import numpy
>>> numpy.oldnumeric.mlab
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'oldnumeric'
  I use Mac machine and did download and install Scipy...

Linda


From aisaac at american.edu  Wed Nov  7 19:57:35 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Wed, 7 Nov 2007 19:57:35 -0500
Subject: [SciPy-user] about MLab
In-Reply-To: <1d987df30711071649j4fd428j4aa999a04c6aea2a@mail.gmail.com>
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com><472DF7C9.8020305@enthought.com><1d987df30711071649j4fd428j4aa999a04c6aea2a@mail.gmail.com>
Message-ID: <Mahogany-0.67.0-952-20071107-195735.00@american.edu>

On Wed, 7 Nov 2007, "linda.s" apparently wrote:
> I still have the problem such as:
> >>> import numpy
> >>> numpy.oldnumeric.mlab 

That is incorrect.
Apparently you missed my post:
>>> from numpy.oldnumeric import mlab

hth,
Alan Isaac





From garyrob at mac.com  Wed Nov  7 20:43:29 2007
From: garyrob at mac.com (Gary Robinson)
Date: Wed, 7 Nov 2007 20:43:29 -0500
Subject: [SciPy-user] genetic algorithms
Message-ID: <20071107204329385576.30c69279@mac.com>

I've seen a few references to a GA module in scipy, as either scipy.ga or sandbox.ga.

But I can't find it in scipy in either of those locations, or elsewhere. (I have version 0.6.)

Is there still such a module? If so, where can I find it?

BTW, while I'm at it, I'll mention that the INSTALL.TXT file for scipy specifies that you need to have 2.3.x or 2.4.x. I'm using it with 2.5 and it seems OK -- does it really need an earlier version or does that file need to be updated?

Thanks,
Gary

-- 

Gary Robinson
CTO
Emergent Music, LLC
garyrob at mac.com
207-942-3463
Company: http://www.goombah.com
Blog:    http://www.garyrobinson.net


From robert.kern at gmail.com  Wed Nov  7 20:46:13 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 07 Nov 2007 19:46:13 -0600
Subject: [SciPy-user] genetic algorithms
In-Reply-To: <20071107204329385576.30c69279@mac.com>
References: <20071107204329385576.30c69279@mac.com>
Message-ID: <47326A65.6070705@gmail.com>

Gary Robinson wrote:
> I've seen a few references to a GA module in scipy, as either scipy.ga or sandbox.ga.
> 
> But I can't find it in scipy in either of those locations, or elsewhere. (I have version 0.6.)
> 
> Is there still such a module? If so, where can I find it?

scipy/sandbox/ga/

Nothing in the sandbox is built by default. See scipy/sandbox/setup.py for
instructions on enabling them. I've done some work on updating ga for numpy, but
I have not tested any of it. If you get it to work, please let us know.

> BTW, while I'm at it, I'll mention that the INSTALL.TXT file for scipy specifies that you need to have 2.3.x or 2.4.x. I'm using it with 2.5 and it seems OK -- does it really need an earlier version or does that file need to be updated?

It should say "at least 2.3".

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From samrobertsmith at gmail.com  Wed Nov  7 21:32:48 2007
From: samrobertsmith at gmail.com (linda.s)
Date: Wed, 7 Nov 2007 19:32:48 -0700
Subject: [SciPy-user] about MLab
In-Reply-To: <Mahogany-0.67.0-952-20071107-195735.00@american.edu>
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com>
	<472DF7C9.8020305@enthought.com>
	<1d987df30711071649j4fd428j4aa999a04c6aea2a@mail.gmail.com>
	<Mahogany-0.67.0-952-20071107-195735.00@american.edu>
Message-ID: <1d987df30711071832n5c877952n71cb9404f91fa489@mail.gmail.com>

On 11/7/07, Alan G Isaac <aisaac at american.edu> wrote:
> On Wed, 7 Nov 2007, "linda.s" apparently wrote:
> > I still have the problem such as:
> > >>> import numpy
> > >>> numpy.oldnumeric.mlab
>
> That is incorrect.
> Apparently you missed my post:
> >>> from numpy.oldnumeric import mlab
>
> hth,
> Alan Isaac
>
Thanks,
Linda


From sinclaird at ukzn.ac.za  Thu Nov  8 01:29:30 2007
From: sinclaird at ukzn.ac.za (Scott Sinclair)
Date: Thu, 08 Nov 2007 08:29:30 +0200
Subject: [SciPy-user] Geostatistics with Scipy?
In-Reply-To: <6946b9500711071237k17d76122ta2dc9e7fe505c0d8@mail.gmail.com>
References: <6946b9500711071237k17d76122ta2dc9e7fe505c0d8@mail.gmail.com>
Message-ID: <4732C867.F934.009F.0@ukzn.ac.za>

Hi,
 
I'd be keen to hear about what you uncover...
 
Options I'm aware of:
 
1. Use the gstat R package via RPy (http://www.gstat.org/).
 
2. Inelegant but effective, install the gstat executable on your system and call using os.system(). This means that you need to write the data points to a file, write an appropriate gstat command file and then read the gstat output files from your Python code. It makes you feel dirty, but means that you can solve your problem easily within a short time frame..
 
3. Wrap gstat for Python. I have a vague plans of trying to do this one day, but don't hold your breath.
 
4. Have a go at wrapping something like the Geostatistics Template Library (http://gstl.sourceforge.net/)
 
5. or try hacking the Python interface to SGeMS (http://sgems.sourceforge.net/), which is based the gstl code. From what I remember, SGeMS embeds the Python interpreter and has some kind of scripting interface, that may (or may not) be easy to call from your Python code.
 
6. Implement your own Kriging algorithm.
 
7. There's also gslib (http://www.gslib.com/) lurking around. I don't recall why I didn't find it particularly interesting.
 
Hope that helps to direct your investigation.
 
Cheers,
Scott

>>> "Mark Bakker" <markbak at gmail.com> 11/7/2007 22:37 >>>
Hello -

I have been looking for some package that can doe Geostatistics with Scipy.
Mostly Kriging of 2D data.

I expect somebody has done something like this before.

Does anybody know of some code somewhere? 

Thanks,

Mark

Please find our Email Disclaimer here: http://www.ukzn.ac.za/disclaimer/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071108/afe0b009/attachment.html>

From w.richert at gmx.net  Thu Nov  8 03:20:45 2007
From: w.richert at gmx.net (Willi Richert)
Date: Thu, 8 Nov 2007 09:20:45 +0100
Subject: [SciPy-user] Where is the border between SciPy and ML algorithms?
Message-ID: <200711080920.45261.w.richert@gmx.net>

Hi,

the scipy.cluster module up to now contains already algorithms for Vector 
Quantization and Kmeans. In the docs it's said that SOMs are following. Now, 
after having followed the mails mentioning Orange and Elefant: Where do we 
draw the line?

I'm asking, because a student of mine has implemented SWIG-Python-wrappers to 
the C++ implementations of diverse decision trees (continuous VFDT, C4.5 and 
some others), which work quite well. We would of course be glad if those 
would be included in standard SciPy. However, I'm not quite sure, what the 
current policy of the core SciPy developers is regarding non-core algorithms.

Thanks in advance for any light on this topic,
wr


From matthieu.brucher at gmail.com  Thu Nov  8 03:26:59 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 8 Nov 2007 09:26:59 +0100
Subject: [SciPy-user] Where is the border between SciPy and ML
	algorithms?
In-Reply-To: <200711080920.45261.w.richert@gmx.net>
References: <200711080920.45261.w.richert@gmx.net>
Message-ID: <e76aa17f0711080026q119b1aeame7b74433f69ddf19@mail.gmail.com>

Hi,

As there are SWIG files and as Scipy does not depend ATM on SWIG, you could
create a scikit with or see with David Cournapeau how it could fit in its
machine learning scikit.

Matthieu

2007/11/8, Willi Richert <w.richert at gmx.net>:
>
> Hi,
>
> the scipy.cluster module up to now contains already algorithms for Vector
> Quantization and Kmeans. In the docs it's said that SOMs are following.
> Now,
> after having followed the mails mentioning Orange and Elefant: Where do we
> draw the line?
>
> I'm asking, because a student of mine has implemented SWIG-Python-wrappers
> to
> the C++ implementations of diverse decision trees (continuous VFDT, C4.5and
> some others), which work quite well. We would of course be glad if those
> would be included in standard SciPy. However, I'm not quite sure, what the
> current policy of the core SciPy developers is regarding non-core
> algorithms.
>
> Thanks in advance for any light on this topic,
> wr
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071108/4eb15516/attachment.html>

From anand.prabhakar.patil at gmail.com  Thu Nov  8 03:46:02 2007
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Thu, 8 Nov 2007 08:46:02 +0000
Subject: [SciPy-user] Geostatistics with Scipy?
Message-ID: <2bc7a5a50711080046u1d3aa1carc3dbbd9365c41516@mail.gmail.com>

Mark Bakker wrote:

>
> Hello -
>
> I have been looking for some package that can doe Geostatistics with Scipy.
> Mostly Kriging of 2D data.
>
> I expect somebody has done something like this before.
>
> Does anybody know of some code somewhere?
>
> Thanks,
>
> Mark

Hi Mark,

You could try the observe() function in my Gaussian process package at
code.google.com/p/random-realizations. Don't tell anyone, I haven't
released it yet. ;) Email me if you try it and need help.

Anand


From david at ar.media.kyoto-u.ac.jp  Thu Nov  8 04:12:19 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Thu, 08 Nov 2007 18:12:19 +0900
Subject: [SciPy-user] Where is the border between SciPy and ML
	algorithms?
In-Reply-To: <200711080920.45261.w.richert@gmx.net>
References: <200711080920.45261.w.richert@gmx.net>
Message-ID: <4732D2F3.3060506@ar.media.kyoto-u.ac.jp>

Willi Richert wrote:
> Hi,
>
> the scipy.cluster module up to now contains already algorithms for Vector 
> Quantization and Kmeans. In the docs it's said that SOMs are following. Now, 
> after having followed the mails mentioning Orange and Elefant: Where do we 
> draw the line?
>   
My understanding is that we do not throw everything we can think of in 
scipy; scipy.cluster was kept  in scipy for backward compatibility ? 
That being said, I don't remember having seen any mention of line 
between core and non core algorithms.

Maybe the main difference between scikits and scipy is the license: 
whereas scipy avoids any non BSD -like license, scikits is freeer in 
this respect (you can depend on GPL code, for example).

> I'm asking, because a student of mine has implemented SWIG-Python-wrappers to 
> the C++ implementations of diverse decision trees (continuous VFDT, C4.5 and 
> some others), which work quite well. We would of course be glad if those 
> would be included in standard SciPy. However, I'm not quite sure, what the 
> current policy of the core SciPy developers is regarding non-core algorithms.
>   
If this is under open source license, as Matthieu said, we (I) can 
import it into the learn scikits, which implements already several 
machine learning algorithms (EM for Gaussian mixtures, datasets, SVM).

cheers,

David


From fredmfp at gmail.com  Thu Nov  8 04:28:46 2007
From: fredmfp at gmail.com (fred)
Date: Thu, 08 Nov 2007 10:28:46 +0100
Subject: [SciPy-user] Geostatistics with Scipy?
In-Reply-To: <6946b9500711071237k17d76122ta2dc9e7fe505c0d8@mail.gmail.com>
References: <6946b9500711071237k17d76122ta2dc9e7fe505c0d8@mail.gmail.com>
Message-ID: <4732D6CE.7000002@gmail.com>

Mark Bakker a ?crit :
> Does anybody know of some code somewhere?
In my own company (http://www.estimages.com),
I develop our kriging codes with scipy+f2py+traits (chaco2 & mayavi2 for 
the 2D & 3D data viz') for the UI.
But my answer is quite useless, as we don't share our python code for now.

Anyway, did you look at this:

http://www.ai-geostats.org

Don't know if there is python code.

Cheers,

-- 
http://scipy.org/FredericPetit



From j.reid at mail.cryst.bbk.ac.uk  Thu Nov  8 04:53:15 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Thu, 08 Nov 2007 09:53:15 +0000
Subject: [SciPy-user] Undefined variable bug in
	scipy.stats.gaussian_kde.integrate_1d
In-Reply-To: <4731FE15.7050503@gmail.com>
References: <fgsnl3$nn9$2@ger.gmane.org> <4731FE15.7050503@gmail.com>
Message-ID: <fgumbl$c5c$1@ger.gmane.org>



Robert Kern wrote:
> When reporting errors, please state the version which you are using. In this
> case, the errors have already been fixed in SVN.
> 

How has it been fixed in SVN? by removing the 'np.' ?



From oliphant at enthought.com  Thu Nov  8 10:15:23 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Thu, 08 Nov 2007 09:15:23 -0600
Subject: [SciPy-user] about MLab
In-Reply-To: <Mahogany-0.67.0-952-20071107-195735.00@american.edu>
References: <1d987df30711040208l53e36652q909252d34ed6fcd1@mail.gmail.com><472DF7C9.8020305@enthought.com><1d987df30711071649j4fd428j4aa999a04c6aea2a@mail.gmail.com>
	<Mahogany-0.67.0-952-20071107-195735.00@american.edu>
Message-ID: <4733280B.6040804@enthought.com>

Alan G Isaac wrote:
> On Wed, 7 Nov 2007, "linda.s" apparently wrote:
>   
>> I still have the problem such as:
>>     
>>>>> import numpy
>>>>> numpy.oldnumeric.mlab 
>>>>>           
>
> That is incorrect.
> Apparently you missed my post:
>   
>>>> from numpy.oldnumeric import mlab
>>>>         
>
>   

You need to import oldnumeric specifically (it is not loaded by default).

import numpy.oldnumeric

Then, you will have

numpy.oldnumeric.mlab


-Travis



From garyrob at mac.com  Thu Nov  8 11:42:41 2007
From: garyrob at mac.com (Gary Robinson)
Date: Thu, 8 Nov 2007 11:42:41 -0500
Subject: [SciPy-user] genetic algorithms
Message-ID: <20071108114241470129.c7dd7bf8@mac.com>

> scipy/sandbox/ga/
> 
> Nothing in the sandbox is built by default. See scipy/sandbox/setup.py for
> instructions on enabling them.

It's not in scipy/sandbox/setup.py. setup.py has the appropriate lines for 'montecarlo', 'spline', and others, but no 'ga'.

Again, I have version = '0.6.0', according to scipy/version.py.

-- 

Gary Robinson
CTO
Emergent Music, LLC
garyrob at mac.com
207-942-3463
Company: http://www.goombah.com
Blog:    http://www.garyrobinson.net


From robert.kern at gmail.com  Thu Nov  8 13:14:03 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 08 Nov 2007 12:14:03 -0600
Subject: [SciPy-user] genetic algorithms
In-Reply-To: <20071108114241470129.c7dd7bf8@mac.com>
References: <20071108114241470129.c7dd7bf8@mac.com>
Message-ID: <473351EB.3070609@gmail.com>

Gary Robinson wrote:
>> scipy/sandbox/ga/
>>
>> Nothing in the sandbox is built by default. See scipy/sandbox/setup.py for
>> instructions on enabling them.
> 
> It's not in scipy/sandbox/setup.py. setup.py has the appropriate lines for 'montecarlo', 'spline', and others, but no 'ga'.

I said to read the instructions in there, namely this:

    # All subpackages should be commented out in the version
    # committed to the repository. This prevents build problems
    # for people who are not actively working with these
    # potentially unstable packages.

    # You can put a list of modules you want to always enable in the
    # file 'enabled_packages.txt' in this directory (you'll have to create it).
    # Since this isn't under version control, it's less likely you'll
    # check it in and screw other people up :-)

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From robert.kern at gmail.com  Thu Nov  8 13:28:06 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 08 Nov 2007 12:28:06 -0600
Subject: [SciPy-user] Undefined variable bug
	in	scipy.stats.gaussian_kde.integrate_1d
In-Reply-To: <fgumbl$c5c$1@ger.gmane.org>
References: <fgsnl3$nn9$2@ger.gmane.org> <4731FE15.7050503@gmail.com>
	<fgumbl$c5c$1@ger.gmane.org>
Message-ID: <47335536.9050000@gmail.com>

John Reid wrote:
> 
> Robert Kern wrote:
>> When reporting errors, please state the version which you are using. In this
>> case, the errors have already been fixed in SVN.
> 
> How has it been fixed in SVN? by removing the 'np.' ?

Yes.

You can find this information yourself, too:

  http://scipy.org/scipy/scipy/browser/trunk/scipy/stats/kde.py

http://scipy.org/scipy/scipy/log/trunk/scipy/stats/kde.py?action=stop_on_copy&rev=3473&stop_rev=&mode=follow_copy

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From garyrob at mac.com  Thu Nov  8 13:32:05 2007
From: garyrob at mac.com (Gary Robinson)
Date: Thu, 8 Nov 2007 13:32:05 -0500
Subject: [SciPy-user] genetic algorithms
In-Reply-To: <473351EB.3070609@gmail.com>
References: <20071108114241470129.c7dd7bf8@mac.com>
	<473351EB.3070609@gmail.com>
Message-ID: <20071108133205031383.c5f07871@mac.com>

Ah. For no good reason I was assuming that if it wasn't one of the packages listed in the commented code in setup.py, it wasn't available.

Thanks.

-- 

Gary Robinson
CTO
Emergent Music, LLC
garyrob at mac.com
207-942-3463
Company: http://www.goombah.com
Blog:    http://www.garyrobinson.net

On Thu, 08 Nov 2007 12:14:03 -0600, Robert Kern wrote:
> Gary Robinson wrote:
>>> scipy/sandbox/ga/
>>> 
>>> Nothing in the sandbox is built by default. See scipy/sandbox/setup.py for
>>> instructions on enabling them.
>> 
>> It's not in scipy/sandbox/setup.py. setup.py has the appropriate 
>> lines for 'montecarlo', 'spline', and others, but no 'ga'.
> 
> I said to read the instructions in there, namely this:
> 
>     # All subpackages should be commented out in the version
>     # committed to the repository. This prevents build problems
>     # for people who are not actively working with these
>     # potentially unstable packages.
> 
>     # You can put a list of modules you want to always enable in the
>     # file 'enabled_packages.txt' in this directory (you'll have to 
> create it).
>     # Since this isn't under version control, it's less likely you'll
>     # check it in and screw other people up :-)


From david.huard at gmail.com  Thu Nov  8 16:33:21 2007
From: david.huard at gmail.com (David Huard)
Date: Thu, 8 Nov 2007 16:33:21 -0500
Subject: [SciPy-user] Argos data
Message-ID: <91cf711d0711081333v8be1cd1y9677243eaf296f75@mail.gmail.com>

Hi,

Is there someone working with Argos satellite data ? I'd need some utility
to process the data files and wanted to know what are the existing tools.


Thanks,

David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071108/652cf20a/attachment.html>

From faltet at carabos.com  Fri Nov  9 04:37:17 2007
From: faltet at carabos.com (Francesc Altet)
Date: Fri, 9 Nov 2007 10:37:17 +0100
Subject: [SciPy-user] Argos data
In-Reply-To: <91cf711d0711081333v8be1cd1y9677243eaf296f75@mail.gmail.com>
References: <91cf711d0711081333v8be1cd1y9677243eaf296f75@mail.gmail.com>
Message-ID: <200711091037.19270.faltet@carabos.com>

A Thursday 08 November 2007, David Huard escrigu?:
> Hi,
>
> Is there someone working with Argos satellite data ? I'd need some
> utility to process the data files and wanted to know what are the
> existing tools.

It would help if you know the kind of format.  My guess is that it 
should be netCDF, but if you can tell us which extension do have these 
files it would be better.  Also, if you are on a Unix system, trying 
the next command out:

$ file your_data_file

should tell you the format of the file (if it is known enough).

In case it is a netCDF file (either v3 or v4), you can use the netCDF4 
python module:

http://netcdf4-python.googlecode.com/svn/trunk/docs/netCDF4-module.html

HTH,

-- 
>0,0<   Francesc Altet ? ? http://www.carabos.com/
V   V   C?rabos Coop. V. ??Enjoy Data
 "-"


From lorenzo.isella at gmail.com  Fri Nov  9 05:08:35 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Fri, 9 Nov 2007 11:08:35 +0100
Subject: [SciPy-user] Introduction to Image Analysis
Message-ID: <a2b3004b0711090208g402fad86u8c520485ecd3fb08@mail.gmail.com>

Dear All,
For my own personal interest, I would like to learn something about
analyzing images with Python.
The number of modules, scripts etc..."out there" can be a bit
confusing, especially if you are starting from scratch.
My purpose is to learn the ropes with some easy-going tutorial (if it
relied on scipy it would be a huge plus for me).
The interest comes from the fact that several job positions I would
like to apply for  imply dealing with satellite images (to detect
common structures, take statistics etc...).
Any suggestions?
Kind Regards

Lorenzo


From schut at sarvision.nl  Fri Nov  9 05:27:44 2007
From: schut at sarvision.nl (Vincent Schut)
Date: Fri, 09 Nov 2007 11:27:44 +0100
Subject: [SciPy-user] Argos data
In-Reply-To: <91cf711d0711081333v8be1cd1y9677243eaf296f75@mail.gmail.com>
References: <91cf711d0711081333v8be1cd1y9677243eaf296f75@mail.gmail.com>
Message-ID: <fh1cn2$e4f$1@ger.gmane.org>

David Huard wrote:
> Hi,
> 
> Is there someone working with Argos satellite data ? I'd need some
> utility to process the data files and wanted to know what are the
> existing tools.

>From quick investigation it is netCDF?
I'd recommend checking out gdal: an open source C library for
reading/writing a multitude of geospatial raster formats, including
stuff like projection info etc. It has a very active community with
great support (have a look at the mailing list archives!), even for
implementing new features/data formats if necessary. And best of all, it
has pretty good python bindings which will allow you to very easy read
(part of) your data into a numpy array.

OK, enough propaganda. Have a look yourself: www.gdal.org.

Cheers,
Vincent.
> 
> 
> Thanks,
> 
> David
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From chanley at stsci.edu  Fri Nov  9 08:29:17 2007
From: chanley at stsci.edu (Christopher Hanley)
Date: Fri, 09 Nov 2007 08:29:17 -0500
Subject: [SciPy-user] Introduction to Image Analysis
In-Reply-To: <a2b3004b0711090208g402fad86u8c520485ecd3fb08@mail.gmail.com>
References: <a2b3004b0711090208g402fad86u8c520485ecd3fb08@mail.gmail.com>
Message-ID: <473460AD.5020804@stsci.edu>

Lorenzo Isella wrote:
> Dear All,
> For my own personal interest, I would like to learn something about
> analyzing images with Python.
> The number of modules, scripts etc..."out there" can be a bit
> confusing, especially if you are starting from scratch.
> My purpose is to learn the ropes with some easy-going tutorial (if it
> relied on scipy it would be a huge plus for me).
> The interest comes from the fact that several job positions I would
> like to apply for  imply dealing with satellite images (to detect
> common structures, take statistics etc...).
> Any suggestions?
> Kind Regards
> 
> Lorenzo
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

Lorenzo,

You could check out this tutorial:

http://www.scipy.org/wikis/topical_software/Tutorial

The examples use astronomical data since it was written by a couple of 
people here at Space Telescope.  However, it is pretty generic.

Chris


-- 
Christopher Hanley
Systems Software Engineer
Space Telescope Science Institute
3700 San Martin Drive
Baltimore MD, 21218
(410) 338-4338


From vel.accel at gmail.com  Fri Nov  9 11:50:22 2007
From: vel.accel at gmail.com (D Hering)
Date: Fri, 9 Nov 2007 11:50:22 -0500
Subject: [SciPy-user] sandbox.timeseries missing cseries
Message-ID: <1e52e0880711090850q4439d944o26b7f3a8f210b214@mail.gmail.com>

Hi.

Just to let you know that sandbox.timeseries __init__.py is missing
"import cseries" and in "__all__ = [...,'cseries']".

dieter


From pgmdevlist at gmail.com  Fri Nov  9 12:12:43 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Fri, 9 Nov 2007 12:12:43 -0500
Subject: [SciPy-user] sandbox.timeseries missing cseries
In-Reply-To: <1e52e0880711090850q4439d944o26b7f3a8f210b214@mail.gmail.com>
References: <1e52e0880711090850q4439d944o26b7f3a8f210b214@mail.gmail.com>
Message-ID: <200711091212.45103.pgmdevlist@gmail.com>

On Friday 09 November 2007 11:50:22 D Hering wrote:
> Hi.
>
> Just to let you know that sandbox.timeseries __init__.py is missing
> "import cseries" and in "__all__ = [...,'cseries']".

Not needed. The function is cseries are already available through dates.py, 
which is loaded in timeseries.__init__


From david.huard at gmail.com  Fri Nov  9 12:49:08 2007
From: david.huard at gmail.com (David Huard)
Date: Fri, 9 Nov 2007 12:49:08 -0500
Subject: [SciPy-user] Argos data
In-Reply-To: <fh1cn2$e4f$1@ger.gmane.org>
References: <91cf711d0711081333v8be1cd1y9677243eaf296f75@mail.gmail.com>
	<fh1cn2$e4f$1@ger.gmane.org>
Message-ID: <91cf711d0711090949w1af50e2cv19e6e4ff07fe28e8@mail.gmail.com>

Thanks guys,

I'm pretty new at this, so I realize I should have mentioned it's data from
an Ice Mass Balance buoy. The files I got are just text files  with lines
describing the satellite pass, then the buoy position and a hexadecimal
number for each sensor (temperature profile in the ice, water and air,
pressure). They are downloaded from argos-system using a telnet script in a
format called PRV,DS (poetic, isn't ir ?) Writing a script to parse all this
is not necessarily complicated, but I'm guessing it has already been done,
though maybe not in python. Something for the dataset scikit, maybe ?


Cheers,

David




2007/11/9, Vincent Schut <schut at sarvision.nl>:

David Huard wrote:
> > Hi,
> >
> > Is there someone working with Argos satellite data ? I'd need some
> > utility to process the data files and wanted to know what are the
> > existing tools.
>
> >From quick investigation it is netCDF?
> I'd recommend checking out gdal: an open source C library for
> reading/writing a multitude of geospatial raster formats, including
> stuff like projection info etc. It has a very active community with
> great support (have a look at the mailing list archives!), even for
> implementing new features/data formats if necessary. And best of all, it
> has pretty good python bindings which will allow you to very easy read
> (part of) your data into a numpy array.
>
> OK, enough propaganda. Have a look yourself: www.gdal.org.
>
> Cheers,
> Vincent.
> >
> >
> > Thanks,
> >
> > David
> >
> >
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071109/4411f530/attachment.html>

From vel.accel at gmail.com  Fri Nov  9 12:54:00 2007
From: vel.accel at gmail.com (dieter h.)
Date: Fri, 9 Nov 2007 12:54:00 -0500
Subject: [SciPy-user] sandbox.timeseries missing cseries
In-Reply-To: <200711091212.45103.pgmdevlist@gmail.com>
References: <1e52e0880711090850q4439d944o26b7f3a8f210b214@mail.gmail.com>
	<200711091212.45103.pgmdevlist@gmail.com>
Message-ID: <1e52e0880711090954k10c694a6w5a9d08a2d6ffbc7c@mail.gmail.com>

On Nov 9, 2007 12:12 PM, Pierre GM <pgmdevlist at gmail.com> wrote:
>
> On Friday 09 November 2007 11:50:22 D Hering wrote:
> > Hi.
> >
> > Just to let you know that sandbox.timeseries __init__.py is missing
> > "import cseries" and in "__all__ = [...,'cseries']".
>
> Not needed. The function is cseries are already available through dates.py,
> which is loaded in timeseries.__init__
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>

Hmm, Ok, but I got a traceback when trying to import timeseries into
the interpreter.

timeseries imports const.py and const.py wanted cseries. After
including cseries (as mentioned above) in __init__.py for the package,
I was able to import timeseries.

-dieter


From zpincus at stanford.edu  Fri Nov  9 14:34:11 2007
From: zpincus at stanford.edu (Zachary Pincus)
Date: Fri, 9 Nov 2007 14:34:11 -0500
Subject: [SciPy-user] KDE question
Message-ID: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>

Hello all,

I'm using scipy.stats.gaussian_kde to estimate the a probability  
density based on some scattered 2D samples. What I would like to do  
eventually is plot a line that surrounds (say) 90% of the probability  
density.

Clearly this line corresponds to some contour of the density function  
(which I can extract out easily enough from KDE density estimates on  
a grid), but I'm not sure how to figure out what exact contour value  
it is that I want.

Given the KDE density estimates on a grid, I could empirically figure  
out which value is closest, but I was wondering if there's a simpler  
closed-form solution... (And apologies if it's somehow obvious; I'm  
feeling a bit foggy-headed about this.)

Best,

Zach



From robert.kern at gmail.com  Fri Nov  9 15:13:02 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 09 Nov 2007 14:13:02 -0600
Subject: [SciPy-user] KDE question
In-Reply-To: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
Message-ID: <4734BF4E.1090102@gmail.com>

Zachary Pincus wrote:
> Hello all,
> 
> I'm using scipy.stats.gaussian_kde to estimate the a probability  
> density based on some scattered 2D samples. What I would like to do  
> eventually is plot a line that surrounds (say) 90% of the probability  
> density.
> 
> Clearly this line corresponds to some contour of the density function  
> (which I can extract out easily enough from KDE density estimates on  
> a grid), but I'm not sure how to figure out what exact contour value  
> it is that I want.
> 
> Given the KDE density estimates on a grid, I could empirically figure  
> out which value is closest, but I was wondering if there's a simpler  
> closed-form solution... (And apologies if it's somehow obvious; I'm  
> feeling a bit foggy-headed about this.)

I don't think there is a closed-form solution for n>1-D. It shouldn't be too
hard to evaluate the KDE on a grid, then search for the value between 0 and the
peak where the integral is around 90%.


from scipy import stats
from scipy.optimize import brentq

k = stats.gaussian_kde(data)
xy = ... # the grid
dxdy = ... # the area represented by each grid point
z = k(xy)
peak = z.max()

# You can't quite do this since the function is discontinuous, and never
# actually equals 0, but you can do a search along these lines.
# Exercise left for the reader.
def f(s):
    mask = z > s
    return 0.9 - (z[mask].sum() * dxdy)

brentq(f, 0, peak)

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From rex at nosyntax.net  Fri Nov  9 15:39:21 2007
From: rex at nosyntax.net (rex)
Date: Fri, 9 Nov 2007 12:39:21 -0800
Subject: [SciPy-user] KDE question
In-Reply-To: <4734BF4E.1090102@gmail.com>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<4734BF4E.1090102@gmail.com>
Message-ID: <20071109203921.GB17405@nosyntax.net>

Robert Kern <robert.kern at gmail.com> [2007-11-09 12:11]:
>from scipy import stats
>from scipy.optimize import brentq

$ python
Python 2.5 (release25-maint, Dec  9 2006, 14:35:53)
[GCC 4.1.2 20061115 (prerelease) (Debian 4.1.1-20)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy
>>> import scipy
>>> from scipy import stats
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python2.5/site-packages/scipy/stats/__init__.py", line 7, in <module>
    from stats import *
  File "/usr/lib/python2.5/site-packages/scipy/stats/stats.py", line 191, in <module>
    import scipy.special as special
  File "/usr/lib/python2.5/site-packages/scipy/special/__init__.py", line 8, in <module>
    from basic import *
  File "/usr/lib/python2.5/site-packages/scipy/special/basic.py", line 8, in <module>
    from _cephes import *
ImportError: /usr/lib/python2.5/site-packages/scipy/special/_cephes.so: undefined symbol: mtherr

This is with numpy-1.04 & scipy 0.60 compiled with icc & ifort
10.1 & using MKL 10.0.

Any hints appreciated (I took your suggestion to drop SUSE ;).

Thanks,

-rex


From zpincus at stanford.edu  Fri Nov  9 16:26:15 2007
From: zpincus at stanford.edu (Zachary Pincus)
Date: Fri, 9 Nov 2007 16:26:15 -0500
Subject: [SciPy-user] KDE question
In-Reply-To: <4734BF4E.1090102@gmail.com>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<4734BF4E.1090102@gmail.com>
Message-ID: <A774FF5E-1602-4AE5-9947-7E009A6840B3@stanford.edu>

Thanks a lot, Robert -- I appreciate your time and suggestions!

Zach

On Nov 9, 2007, at 3:13 PM, Robert Kern wrote:

> Zachary Pincus wrote:
>> Hello all,
>>
>> I'm using scipy.stats.gaussian_kde to estimate the a probability
>> density based on some scattered 2D samples. What I would like to do
>> eventually is plot a line that surrounds (say) 90% of the probability
>> density.
>>
>> Clearly this line corresponds to some contour of the density function
>> (which I can extract out easily enough from KDE density estimates on
>> a grid), but I'm not sure how to figure out what exact contour value
>> it is that I want.
>>
>> Given the KDE density estimates on a grid, I could empirically figure
>> out which value is closest, but I was wondering if there's a simpler
>> closed-form solution... (And apologies if it's somehow obvious; I'm
>> feeling a bit foggy-headed about this.)
>
> I don't think there is a closed-form solution for n>1-D. It  
> shouldn't be too
> hard to evaluate the KDE on a grid, then search for the value  
> between 0 and the
> peak where the integral is around 90%.
>
>
> from scipy import stats
> from scipy.optimize import brentq
>
> k = stats.gaussian_kde(data)
> xy = ... # the grid
> dxdy = ... # the area represented by each grid point
> z = k(xy)
> peak = z.max()
>
> # You can't quite do this since the function is discontinuous, and  
> never
> # actually equals 0, but you can do a search along these lines.
> # Exercise left for the reader.
> def f(s):
>     mask = z > s
>     return 0.9 - (z[mask].sum() * dxdy)
>
> brentq(f, 0, peak)
>
> -- 
> Robert Kern
>
> "I have come to believe that the whole world is an enigma, a  
> harmless enigma
>  that is made terrible by our own mad attempt to interpret it as  
> though it had
>  an underlying truth."
>   -- Umberto Eco
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From yarlett at psych.stanford.edu  Fri Nov  9 16:33:23 2007
From: yarlett at psych.stanford.edu (Daniel Yarlett)
Date: Fri, 9 Nov 2007 13:33:23 -0800
Subject: [SciPy-user] Problem installing scipy 0.6.0 on PPC OS X 10.5
	Leopard.
Message-ID: <16DF925E-0844-48C5-806D-E9F00488D862@psych.stanford.edu>

hi,

apologies if this is not the appropriate forum for posting this, but i  
am having some trouble building and installing scipy 0.6.0 on a PPC  
Mac that i have just upgraded to OS X 10.5 (this version of scipy  
previously worked fine on the same machine under OS X 10.4).

numpy 1.0.4 seems to install fine with "python setup.py install".

however, when i try to install scipy 0.6.0 i get the output (tail  
only) included at the end of this email.

here are my compiler versions:

gcc -- version
gcc (GCC) 4.3.0 20071026 (experimental)

gfortran -- version
GNU Fortran (GCC) 4.3.0 20071026 (experimental)

i remember that on previous versions of OS X you could use gcc_select  
to specify v3.4 of gcc to make the build go through. unfortunately  
gcc_select seems to have been deprecated in Leopard so i haven't been  
able to try that. don't know if that's a red herring or not.

also, fwiw, i just upgraded an intel macbook prop from OS X 10.4 to  
10.5 and had no problems at all with the pre-existing versions of  
numpy and scipy.

anyway, any help would be greatly appreciated!

best,

dan.

--

END OF OUTPUT FROM "python setup.py install" for scipy 0.6.0:

Could not locate executable f95
customize AbsoftFCompiler
Could not locate executable f90
Could not locate executable f77
customize IBMFCompiler
Could not locate executable xlf90
Could not locate executable xlf
customize IntelFCompiler
Could not locate executable ifort
Could not locate executable ifc
customize GnuFCompiler
Found executable /usr/local/bin/g77
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler using build_clib
building 'superlu_src' library
compiling C sources
C compiler: gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/ 
MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp-precomp - 
mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3

compile options: '-DUSE_VENDOR_BLAS=1 -c'
gcc: scipy/linsolve/SuperLU/SRC/colamd.c
gcc: unrecognized option '-no-cpp-precomp'
cc1: error: unrecognized command line option "-arch"
cc1: error: unrecognized command line option "-arch"
cc1: error: unrecognized command line option "-Wno-long-double"
gcc: unrecognized option '-no-cpp-precomp'
cc1: error: unrecognized command line option "-arch"
cc1: error: unrecognized command line option "-arch"
cc1: error: unrecognized command line option "-Wno-long-double"
error: Command "gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/ 
MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp-precomp - 
mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3 - 
DUSE_VENDOR_BLAS=1 -c scipy/linsolve/SuperLU/SRC/colamd.c -o build/ 
temp.macosx-10.3-fat-2.5/scipy/linsolve/SuperLU/SRC/colamd.o" failed  
with exit status 1


From ellisonbg.net at gmail.com  Fri Nov  9 16:47:29 2007
From: ellisonbg.net at gmail.com (Brian Granger)
Date: Fri, 9 Nov 2007 14:47:29 -0700
Subject: [SciPy-user] Problem installing scipy 0.6.0 on PPC OS X 10.5
	Leopard.
In-Reply-To: <16DF925E-0844-48C5-806D-E9F00488D862@psych.stanford.edu>
References: <16DF925E-0844-48C5-806D-E9F00488D862@psych.stanford.edu>
Message-ID: <6ce0ac130711091347q4dbab44akce022d6b9bf2c72c@mail.gmail.com>

Are you using the builtin python?  If so, you will need to set your
PYTHONPATH to point to the directory where numpy 1.0.4 is installed.
The reason is that the builtin python comes with python 1.0.1, which
won't build scipy with the gfortran you have.

There is a somewhat odd issue on Leopard.  If you look at sys.path:

python -c "import sys; print sys.path"

You will see that the directory where the builtin numpy is (I think it
is something like */Extras/*) is before the site-packages directory
where your numpy 1.0.4 would be installed (unless you used to --home
or --prefix options when running python setup.py install).  This means
that unless you set PYTHONPATH you will always get Leopards older
numpy, which will cause scipy to fail on building.

Note, this is very different than on Tiger where the builtin python
did not include many third party packages.  Leopard includes quite a
few (for better or worse).

Brian



On Nov 9, 2007 2:33 PM, Daniel Yarlett <yarlett at psych.stanford.edu> wrote:
> hi,
>
> apologies if this is not the appropriate forum for posting this, but i
> am having some trouble building and installing scipy 0.6.0 on a PPC
> Mac that i have just upgraded to OS X 10.5 (this version of scipy
> previously worked fine on the same machine under OS X 10.4).
>
> numpy 1.0.4 seems to install fine with "python setup.py install".
>
> however, when i try to install scipy 0.6.0 i get the output (tail
> only) included at the end of this email.
>
> here are my compiler versions:
>
> gcc -- version
> gcc (GCC) 4.3.0 20071026 (experimental)
>
> gfortran -- version
> GNU Fortran (GCC) 4.3.0 20071026 (experimental)
>
> i remember that on previous versions of OS X you could use gcc_select
> to specify v3.4 of gcc to make the build go through. unfortunately
> gcc_select seems to have been deprecated in Leopard so i haven't been
> able to try that. don't know if that's a red herring or not.
>
> also, fwiw, i just upgraded an intel macbook prop from OS X 10.4 to
> 10.5 and had no problems at all with the pre-existing versions of
> numpy and scipy.
>
> anyway, any help would be greatly appreciated!
>
> best,
>
> dan.
>
> --
>
> END OF OUTPUT FROM "python setup.py install" for scipy 0.6.0:
>
> Could not locate executable f95
> customize AbsoftFCompiler
> Could not locate executable f90
> Could not locate executable f77
> customize IBMFCompiler
> Could not locate executable xlf90
> Could not locate executable xlf
> customize IntelFCompiler
> Could not locate executable ifort
> Could not locate executable ifc
> customize GnuFCompiler
> Found executable /usr/local/bin/g77
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler using build_clib
> building 'superlu_src' library
> compiling C sources
> C compiler: gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/
> MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp-precomp -
> mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3
>
> compile options: '-DUSE_VENDOR_BLAS=1 -c'
> gcc: scipy/linsolve/SuperLU/SRC/colamd.c
> gcc: unrecognized option '-no-cpp-precomp'
> cc1: error: unrecognized command line option "-arch"
> cc1: error: unrecognized command line option "-arch"
> cc1: error: unrecognized command line option "-Wno-long-double"
> gcc: unrecognized option '-no-cpp-precomp'
> cc1: error: unrecognized command line option "-arch"
> cc1: error: unrecognized command line option "-arch"
> cc1: error: unrecognized command line option "-Wno-long-double"
> error: Command "gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/
> MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp-precomp -
> mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3 -
> DUSE_VENDOR_BLAS=1 -c scipy/linsolve/SuperLU/SRC/colamd.c -o build/
> temp.macosx-10.3-fat-2.5/scipy/linsolve/SuperLU/SRC/colamd.o" failed
> with exit status 1
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From yarlett at psych.stanford.edu  Fri Nov  9 17:20:02 2007
From: yarlett at psych.stanford.edu (Daniel Yarlett)
Date: Fri, 9 Nov 2007 14:20:02 -0800
Subject: [SciPy-user] Problem installing scipy 0.6.0 on PPC OS X 10.5
	Leopard.
Message-ID: <9EFF3A63-E69C-4C3F-B9AB-9DB36BBD4E21@psych.stanford.edu>

here's what i get when i type "python -c "import sys; print sys.path":
['', '/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python25.zip', '/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5', '/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/plat-darwin', '/Library/Frameworks/Python.framework/Versions/ 
2.5/lib/python2.5/plat-mac', '/Library/Frameworks/Python.framework/ 
Versions/2.5/lib/python2.5/plat-mac/lib-scriptpackages', '/Library/ 
Frameworks/Python.framework/Versions/2.5/lib/python2.5/lib-tk', '/ 
Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/lib- 
dynload', '/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages']
i don't see anything of the form */Extras/* in there, although site- 
packages is the last in the path.

but when i run my default version of python and import numpy i get  
version 1.0.4, which is the latest one i installed, and probably isn't  
the default version that comes with leopard:

ramscar-lab5:~ yarlett$ python -c "import numpy; print  
numpy.__version__"
1.0.4

i can try setting $PYTHONPATH if you still think that might help. if  
so i'd appreciate it if you could let me know what i should set it to.

thanks for the speedy response!

dan.


> Are you using the builtin python?  If so, you will need to set your
> PYTHONPATH to point to the directory where numpy 1.0.4 is installed.
> The reason is that the builtin python comes with python 1.0.1, which
> won't build scipy with the gfortran you have.
>
> There is a somewhat odd issue on Leopard.  If you look at sys.path:
>
> python -c "import sys; print sys.path"
>
> You will see that the directory where the builtin numpy is (I think it
> is something like */Extras/*) is before the site-packages directory
> where your numpy 1.0.4 would be installed (unless you used to --home
> or --prefix options when running python setup.py install).  This means
> that unless you set PYTHONPATH you will always get Leopards older
> numpy, which will cause scipy to fail on building.
>
> Note, this is very different than on Tiger where the builtin python
> did not include many third party packages.  Leopard includes quite a
> few (for better or worse).
>
> Brian
>
> On Nov 9, 2007 2:33 PM, Daniel Yarlett <yarlett <at>  
> psych.stanford.edu> wrote:
> > hi,
> >
> > apologies if this is not the appropriate forum for posting this,  
> but i
> > am having some trouble building and installing scipy 0.6.0 on a PPC
> > Mac that i have just upgraded to OS X 10.5 (this version of scipy
> > previously worked fine on the same machine under OS X 10.4).
> >
> > numpy 1.0.4 seems to install fine with "python setup.py install".
> >
> > however, when i try to install scipy 0.6.0 i get the output (tail
> > only) included at the end of this email.
> >
> > here are my compiler versions:
> >
> > gcc -- version
> > gcc (GCC) 4.3.0 20071026 (experimental)
> >
> > gfortran -- version
> > GNU Fortran (GCC) 4.3.0 20071026 (experimental)
> >
> > i remember that on previous versions of OS X you could use  
> gcc_select
> > to specify v3.4 of gcc to make the build go through. unfortunately
> > gcc_select seems to have been deprecated in Leopard so i haven't  
> been
> > able to try that. don't know if that's a red herring or not.
> >
> > also, fwiw, i just upgraded an intel macbook prop from OS X 10.4 to
> > 10.5 and had no problems at all with the pre-existing versions of
> > numpy and scipy.
> >
> > anyway, any help would be greatly appreciated!
> >
> > best,
> >
> > dan.
> >
> > --
> >
> > END OF OUTPUT FROM "python setup.py install" for scipy 0.6.0:
> >
> > Could not locate executable f95
> > customize AbsoftFCompiler
> > Could not locate executable f90
> > Could not locate executable f77
> > customize IBMFCompiler
> > Could not locate executable xlf90
> > Could not locate executable xlf
> > customize IntelFCompiler
> > Could not locate executable ifort
> > Could not locate executable ifc
> > customize GnuFCompiler
> > Found executable /usr/local/bin/g77
> > gnu: no Fortran 90 compiler found
> > gnu: no Fortran 90 compiler found
> > customize GnuFCompiler
> > gnu: no Fortran 90 compiler found
> > gnu: no Fortran 90 compiler found
> > customize GnuFCompiler using build_clib
> > building 'superlu_src' library
> > compiling C sources
> > C compiler: gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/
> > MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp- 
> precomp -
> > mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3
> >
> > compile options: '-DUSE_VENDOR_BLAS=1 -c'
> > gcc: scipy/linsolve/SuperLU/SRC/colamd.c
> > gcc: unrecognized option '-no-cpp-precomp'
> > cc1: error: unrecognized command line option "-arch"
> > cc1: error: unrecognized command line option "-arch"
> > cc1: error: unrecognized command line option "-Wno-long-double"
> > gcc: unrecognized option '-no-cpp-precomp'
> > cc1: error: unrecognized command line option "-arch"
> > cc1: error: unrecognized command line option "-arch"
> > cc1: error: unrecognized command line option "-Wno-long-double"
> > error: Command "gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/
> > MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp- 
> precomp -
> > mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3 -
> > DUSE_VENDOR_BLAS=1 -c scipy/linsolve/SuperLU/SRC/colamd.c -o build/
> > temp.macosx-10.3-fat-2.5/scipy/linsolve/SuperLU/SRC/colamd.o" failed
> > with exit status 1
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user <at> scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071109/889ca842/attachment.html>

From cckidder at gmail.com  Fri Nov  9 18:49:37 2007
From: cckidder at gmail.com (Chad Kidder)
Date: Fri, 9 Nov 2007 17:49:37 -0600
Subject: [SciPy-user] Submitting new code
Message-ID: <AFC7072E-AD54-4A31-91FB-8DDF9EC2A025@gmail.com>

I've got some file import classes that I am working on and would  
eventually like to contribute back to scipy, probably the io  
package.  I've done a little bit of searching and didn't find  
anything really descriptive.  Would anyone like to give me / point to  
a bit of information on how to go about this?


--Chad Kidder


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071109/f1940665/attachment.html>

From oliphant at enthought.com  Fri Nov  9 19:01:13 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Fri, 09 Nov 2007 18:01:13 -0600
Subject: [SciPy-user] Submitting new code
In-Reply-To: <AFC7072E-AD54-4A31-91FB-8DDF9EC2A025@gmail.com>
References: <AFC7072E-AD54-4A31-91FB-8DDF9EC2A025@gmail.com>
Message-ID: <4734F4C9.10507@enthought.com>

Chad Kidder wrote:
> I've got some file import classes that I am working on and would 
> eventually like to contribute back to scipy, probably the io package.  
> I've done a little bit of searching and didn't find anything really 
> descriptive.  Would anyone like to give me / point to a bit of 
> information on how to go about this? 
>
>
Create a ticket at the Trac page for SciPy and upload a patch with your 
new code in it.

This is the easiest way.   Ask if you need more help.

Good luck,

-Travis O.


>
> --Chad Kidder
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From roger.herikstad at gmail.com  Sat Nov 10 01:16:06 2007
From: roger.herikstad at gmail.com (Roger Herikstad)
Date: Sat, 10 Nov 2007 14:16:06 +0800
Subject: [SciPy-user] Inverse of complex matrix
Message-ID: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>

Dear scipy users, I am trying to invert a 12 by 12 complex matrix R using
scipy.linalg.inv(), but the round-off errors seem a bit too big. If I do
scipy.dot(Ri,R), the first row of the resulting matrix looks like this:

array([ 0.953125  +0.j        ,  0.06542969+0.01171875j,
        0.08154297+0.15234375j, -0.078125  -0.09375j   ,
        0.19140625-0.0859375j ,  0.08203125+0.140625j  ,
       -0.08398438+0.0078125j , -0.02539062-0.03125j   ,
       -0.06738281+0.015625j  , -0.04736328+0.15820312j,
       -0.09057617+0.02832031j,  0.109375  -0.0625j    ])


Is this normal? Inverting a 2x2 matrix produces the expected result. I've
attached a pickle of my matrix. Thanks!

~ Roger
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071110/d2c6e10c/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: matrix.pck
Type: application/octet-stream
Size: 7072 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071110/d2c6e10c/attachment.obj>

From nwagner at iam.uni-stuttgart.de  Sat Nov 10 02:54:17 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sat, 10 Nov 2007 08:54:17 +0100
Subject: [SciPy-user] Inverse of complex matrix
In-Reply-To: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>
References: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>
Message-ID: <web-91832783@uni-stuttgart.de>

On Sat, 10 Nov 2007 14:16:06 +0800
  "Roger Herikstad" <roger.herikstad at gmail.com> wrote:
> Dear scipy users, I am trying to invert a 12 by 12 
>complex matrix R using
> scipy.linalg.inv(), but the round-off errors seem a bit 
>too big. If I do
> scipy.dot(Ri,R), the first row of the resulting matrix 
>looks like this:
> 
> array([ 0.953125  +0.j        ,  0.06542969+0.01171875j,
>        0.08154297+0.15234375j, -0.078125  -0.09375j   ,
>        0.19140625-0.0859375j ,  0.08203125+0.140625j  ,
>       -0.08398438+0.0078125j , -0.02539062-0.03125j   ,
>       -0.06738281+0.015625j  , -0.04736328+0.15820312j,
>       -0.09057617+0.02832031j,  0.109375  -0.0625j    ])
> 
> 
> Is this normal? Inverting a 2x2 matrix produces the 
>expected result. I've
> attached a pickle of my matrix. Thanks!
> 
> ~ Roger
  
It depends on the condition number of your matrix.
Did you compute the singular values of the matrix ?

Please can you send me your test matrix in matrixmarket
format off-list.
io.mmwrite

Nils


From mforbes at physics.ubc.ca  Sat Nov 10 02:59:43 2007
From: mforbes at physics.ubc.ca (Michael McNeil Forbes)
Date: Fri, 9 Nov 2007 23:59:43 -0800
Subject: [SciPy-user] Inverse of complex matrix
In-Reply-To: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>
References: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>
Message-ID: <7141ED62-C305-41CB-AB7B-703C01C74F72@physics.ubc.ca>

Hi Roger,

Your matrix is extremely singular.  The singular values are:

 >>> U,d,V = linalg.svd(R)
 >>> d
array([1.41944451e+01,   1.23913727e+01,   9.46227279e+00,
          2.81247753e-04,   1.11035378e-14,   1.02548448e-14,
          8.22627497e-15,   6.74668324e-15,   5.51882714e-15,
          4.78886237e-15,   2.82020737e-15,   7.14721524e-16])

Thus, the matrix is really il-conditioned (the ratio of the largest  
to smallest singular value is of the order of the machine precision.   
It is essentially singular, and only has 4 linearly independent  
columns).   Thus, it is not surprising that the round-off errors are  
significant.

You can't reasonably expect to invert a matrix when the ratio of the  
singular values is of order the machine precision.

Michael.

On 9 Nov 2007, at 10:16 PM, Roger Herikstad wrote:

> Dear scipy users,
>  I am trying to invert a 12 by 12 complex matrix R using  
> scipy.linalg.inv(), but the round-off errors seem a bit too big. If  
> I do scipy.dot(Ri,R), the first row of the resulting matrix looks  
> like this:
>
> array([ 0.953125  +0.j        ,  0.06542969+0.01171875j,
>         0.08154297+0.15234375j, -0.078125  -0.09375j   ,
>         0.19140625-0.0859375j ,  0.08203125+0.140625j  ,
>        -0.08398438+0.0078125j , -0.02539062-0.03125j   ,
>        -0.06738281+0.015625j  , -0.04736328+0.15820312j,
>        -0.09057617+0.02832031j,  0.109375  -0.0625j    ])
>
>
> Is this normal? Inverting a 2x2 matrix produces the expected  
> result. I've attached a pickle of my matrix. Thanks!
>
> ~ Roger
> <matrix.pck>
>



From david at ar.media.kyoto-u.ac.jp  Sat Nov 10 02:53:06 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Sat, 10 Nov 2007 16:53:06 +0900
Subject: [SciPy-user] KDE question
In-Reply-To: <4734BF4E.1090102@gmail.com>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<4734BF4E.1090102@gmail.com>
Message-ID: <47356362.7010508@ar.media.kyoto-u.ac.jp>

Robert Kern wrote:
> Zachary Pincus wrote:
>> Hello all,
>>
>> I'm using scipy.stats.gaussian_kde to estimate the a probability  
>> density based on some scattered 2D samples. What I would like to do  
>> eventually is plot a line that surrounds (say) 90% of the probability  
>> density.
>>
>> Clearly this line corresponds to some contour of the density function  
>> (which I can extract out easily enough from KDE density estimates on  
>> a grid), but I'm not sure how to figure out what exact contour value  
>> it is that I want.
>>
>> Given the KDE density estimates on a grid, I could empirically figure  
>> out which value is closest, but I was wondering if there's a simpler  
>> closed-form solution... (And apologies if it's somehow obvious; I'm  
>> feeling a bit foggy-headed about this.)
>
> I don't think there is a closed-form solution for n>1-D. It shouldn't be too
> hard to evaluate the KDE on a grid, then search for the value between 0 and the
> peak where the integral is around 90%.
I am not sure I understand exactly the problem, but if the problem is to 
find a contour level of a Gaussian density, it has a closed form for any 
dimension. For the actual code, you can look at 
scikits.learn.machine.em, function gauss_ell, which is used by my Em 
toolbox for gaussian mixtures

http://projects.scipy.org/scipy/scikits/browser/trunk/learn/scikits/learn/machine/em/densities.py

The function can be used without any other code, I think (at least, I 
remember I wanted it that way:) ). This is based on the fact that if X ~ 
N(mu, sigma), then (X-mu)' * sigma^-1 * (X-mu) is a one dimensional 
random variable following a chi square.

cheers,

David


From robert.kern at gmail.com  Sat Nov 10 03:08:04 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Sat, 10 Nov 2007 02:08:04 -0600
Subject: [SciPy-user] KDE question
In-Reply-To: <47356362.7010508@ar.media.kyoto-u.ac.jp>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>	<4734BF4E.1090102@gmail.com>
	<47356362.7010508@ar.media.kyoto-u.ac.jp>
Message-ID: <473566E4.2060101@gmail.com>

David Cournapeau wrote:

> I am not sure I understand exactly the problem, but if the problem is to 
> find a contour level of a Gaussian density, it has a closed form for any 
> dimension.

No, the problem is to find the appropriate contour level of a kernel density
estimate (with Gaussian kernels in this case). Essentially, a mixture of many
Gaussians, not a single one.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From david at ar.media.kyoto-u.ac.jp  Sat Nov 10 03:11:05 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Sat, 10 Nov 2007 17:11:05 +0900
Subject: [SciPy-user] Inverse of complex matrix
In-Reply-To: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>
References: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>
Message-ID: <47356799.8000109@ar.media.kyoto-u.ac.jp>

Roger Herikstad wrote:
> Dear scipy users,
>  I am trying to invert a 12 by 12 complex matrix R using 
> scipy.linalg.inv(), but the round-off errors seem a bit too big. If I 
> do scipy.dot(Ri,R), the first row of the resulting matrix looks like 
> this:
>
> array([ 0.953125  +0.j        ,  0.06542969+0.01171875j,
>         0.08154297+0.15234375j, -0.078125  -0.09375j   ,
>         0.19140625-0.0859375j ,  0.08203125+0.140625j  ,
>        -0.08398438+0.0078125j , -0.02539062-0.03125j   ,
>        -0.06738281+0.015625j  , -0.04736328+0.15820312j,
>        -0.09057617+0.02832031j,  0.109375  -0.0625j    ])
>
>  
> Is this normal? Inverting a 2x2 matrix produces the expected result. 
> I've attached a pickle of my matrix. Thanks!
As a rule, you should not invert a matrix unless you *really* need it. 
For example, if you want X in
A X = B, with A a matrix, B and X vectors, inverting A should be 
avoided. Some other methods are better suited (numerically speaking), 
depending on A and what you are trying to do.

The mathworks have good introduction on this topic, if you look for 
online help on inv, pinv and so on:

http://www.mathworks.com/access/helpdesk/help/techdoc/ref/inv.html

cheers,

David



From david at ar.media.kyoto-u.ac.jp  Sat Nov 10 03:24:18 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Sat, 10 Nov 2007 17:24:18 +0900
Subject: [SciPy-user] KDE question
In-Reply-To: <473566E4.2060101@gmail.com>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>	<4734BF4E.1090102@gmail.com>	<47356362.7010508@ar.media.kyoto-u.ac.jp>
	<473566E4.2060101@gmail.com>
Message-ID: <47356AB2.2030009@ar.media.kyoto-u.ac.jp>

Robert Kern wrote:
> David Cournapeau wrote:
>
>> I am not sure I understand exactly the problem, but if the problem is to 
>> find a contour level of a Gaussian density, it has a closed form for any 
>> dimension.
>
> No, the problem is to find the appropriate contour level of a kernel density
> estimate (with Gaussian kernels in this case). Essentially, a mixture of many
> Gaussians, not a single one.
>
Ah ok, sorry.

David


From roger.herikstad at gmail.com  Sat Nov 10 05:09:20 2007
From: roger.herikstad at gmail.com (Roger Herikstad)
Date: Sat, 10 Nov 2007 18:09:20 +0800
Subject: [SciPy-user] Inverse of complex matrix
In-Reply-To: <47356799.8000109@ar.media.kyoto-u.ac.jp>
References: <bae941750711092216o60036566n986a813ce3d525e@mail.gmail.com>
	<47356799.8000109@ar.media.kyoto-u.ac.jp>
Message-ID: <bae941750711100209n12dc3aey1a074b46a266911a@mail.gmail.com>

Dear all,
 Thanks. I can get away with not inverting the matrix, actually, I
just wanted to confirm a calculation via a different route, and I was
momentarily surprised by my result. I realise my mistake though.
Thanks for pointing it out for me!

~ Roger

On Nov 10, 2007 4:11 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
>
> Roger Herikstad wrote:
> > Dear scipy users,
> >  I am trying to invert a 12 by 12 complex matrix R using
> > scipy.linalg.inv(), but the round-off errors seem a bit too big. If I
> > do scipy.dot(Ri,R), the first row of the resulting matrix looks like
> > this:
> >
> > array([ 0.953125  +0.j        ,  0.06542969+0.01171875j,
> >         0.08154297+0.15234375j, -0.078125  -0.09375j   ,
> >         0.19140625-0.0859375j ,  0.08203125+0.140625j  ,
> >        -0.08398438+0.0078125j , -0.02539062-0.03125j   ,
> >        -0.06738281+0.015625j  , -0.04736328+0.15820312j,
> >        -0.09057617+0.02832031j,  0.109375  -0.0625j    ])
> >
> >
> > Is this normal? Inverting a 2x2 matrix produces the expected result.
> > I've attached a pickle of my matrix. Thanks!
> As a rule, you should not invert a matrix unless you *really* need it.
> For example, if you want X in
> A X = B, with A a matrix, B and X vectors, inverting A should be
> avoided. Some other methods are better suited (numerically speaking),
> depending on A and what you are trying to do.
>
> The mathworks have good introduction on this topic, if you look for
> online help on inv, pinv and so on:
>
> http://www.mathworks.com/access/helpdesk/help/techdoc/ref/inv.html
>
> cheers,
>
> David
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From gael.varoquaux at normalesup.org  Sat Nov 10 05:24:45 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sat, 10 Nov 2007 11:24:45 +0100
Subject: [SciPy-user] KDE question
In-Reply-To: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
Message-ID: <20071110102445.GB22735@clipper.ens.fr>

On Fri, Nov 09, 2007 at 02:34:11PM -0500, Zachary Pincus wrote:
> I'm using scipy.stats.gaussian_kde to estimate the a probability  
> density based on some scattered 2D samples. What I would like to do  
> eventually is plot a line that surrounds (say) 90% of the probability  
> density.

If its only plotting you are interested in, matplotlib can do all the
dirty work for you, using the "contour" command:

http://matplotlib.sourceforge.net/screenshots/pcolor_demo.py
which yields:
http://matplotlib.sourceforge.net/screenshots/pcolor_demo_large.png

HTH,

Ga?l


From david at ar.media.kyoto-u.ac.jp  Sat Nov 10 06:13:31 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Sat, 10 Nov 2007 20:13:31 +0900
Subject: [SciPy-user] KDE question
In-Reply-To: <20071110102445.GB22735@clipper.ens.fr>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<20071110102445.GB22735@clipper.ens.fr>
Message-ID: <4735925B.6080409@ar.media.kyoto-u.ac.jp>

Gael Varoquaux wrote:
> On Fri, Nov 09, 2007 at 02:34:11PM -0500, Zachary Pincus wrote:
>> I'm using scipy.stats.gaussian_kde to estimate the a probability  
>> density based on some scattered 2D samples. What I would like to do  
>> eventually is plot a line that surrounds (say) 90% of the probability  
>> density.
>
> If its only plotting you are interested in, matplotlib can do all the
> dirty work for you, using the "contour" command:
>
> http://matplotlib.sourceforge.net/screenshots/pcolor_demo.py
> which yields:
> http://matplotlib.sourceforge.net/screenshots/pcolor_demo_large.png
I missed this was for plotting. Then the code from the EM toolbox may 
actually be useful, since I used the function mentioned before to plot 
density contours for mixture of Gaussians:

http://www.ar.media.kyoto-u.ac.jp/members/david/softwares/em/pdfestimation.png

Code there: 
http://projects.scipy.org/scipy/scikits/browser/trunk/learn/scikits/learn/machine/em/gauss_mix.py 
(function density_on_grid: this can be used from a Gaussian mixture 
class, that you may be able use easily from gauss_kde). This may not be 
adapted for your case, since I only coded it with a few Gaussians in 
mind (e.g. I have never tried to do something smart, since I don't need 
mixtures of say 500 Gaussian).

cheers,

David


From zpincus at stanford.edu  Sat Nov 10 07:14:39 2007
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sat, 10 Nov 2007 07:14:39 -0500
Subject: [SciPy-user] KDE question
In-Reply-To: <20071110102445.GB22735@clipper.ens.fr>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<20071110102445.GB22735@clipper.ens.fr>
Message-ID: <9F4EB7F4-D2E5-4363-B93F-105A76501D4F@stanford.edu>

>> I'm using scipy.stats.gaussian_kde to estimate the a probability
>> density based on some scattered 2D samples. What I would like to do
>> eventually is plot a line that surrounds (say) 90% of the probability
>> density.
>
> If its only plotting you are interested in, matplotlib can do all the
> dirty work for you, using the "contour" command:
>
> http://matplotlib.sourceforge.net/screenshots/pcolor_demo.py
> which yields:
> http://matplotlib.sourceforge.net/screenshots/pcolor_demo_large.png

In this case, the trick is to figure out *which* contour level to  
plot -- that is, at what density level and above is 90% (say) of the  
total density contained. Robert's example shows how to figure that  
out, and pretty cleverly. I wouldn't have thought of wrapping it all  
up with brentq.

But gosh, do those matplotlib contours look nice. I've been using  
Gnuplot/Gnuplot.py for a long time, but it might be time to make a  
switch.

Thanks,
Zach


From zpincus at stanford.edu  Sat Nov 10 07:24:31 2007
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sat, 10 Nov 2007 07:24:31 -0500
Subject: [SciPy-user] Problem installing scipy 0.6.0 on PPC OS X 10.5
	Leopard.
In-Reply-To: <6ce0ac130711091347q4dbab44akce022d6b9bf2c72c@mail.gmail.com>
References: <16DF925E-0844-48C5-806D-E9F00488D862@psych.stanford.edu>
	<6ce0ac130711091347q4dbab44akce022d6b9bf2c72c@mail.gmail.com>
Message-ID: <87CE1249-D53F-4B49-8D8A-B6CA182B2D83@stanford.edu>

Hello all,

> Are you using the builtin python?  If so, you will need to set your
> PYTHONPATH to point to the directory where numpy 1.0.4 is installed.
> The reason is that the builtin python comes with python 1.0.1, which
> won't build scipy with the gfortran you have.

I'm not using Leopard, but if I recall from the discussion on python- 
mac, there's the added difficulty that the environment isn't  
preserved through 'sudo' -- so just setting the PYTHONPATH  
environment variable won't work if you need to 'sudo python setup.py  
install' to install a given library to a system directory.

I don't recall what the recommended solution, if any, was. I think it  
involved some .pth file trickery that looked relatively, um, ugly.  
Alternately, one could select a non-system directory as your local  
site-packages, via a .pth or PYTHONPATH, so sudo is not needed to  
install. This doesn't work for system admins supporting multiple  
users though. Finally, on the horizon were some options for setting  
up "virtual" python installs that would act like separate installs  
from the system one for the purpose of installing packages, but would  
use the system python and its libs under the covers.

Is there yet any set of simple, newbie-friendly instructions that  
covers how best to install things like numpy/scipy on leopard?

Zach


On Nov 9, 2007, at 4:47 PM, Brian Granger wrote:

> Are you using the builtin python?  If so, you will need to set your
> PYTHONPATH to point to the directory where numpy 1.0.4 is installed.
> The reason is that the builtin python comes with python 1.0.1, which
> won't build scipy with the gfortran you have.
>
> There is a somewhat odd issue on Leopard.  If you look at sys.path:
>
> python -c "import sys; print sys.path"
>
> You will see that the directory where the builtin numpy is (I think it
> is something like */Extras/*) is before the site-packages directory
> where your numpy 1.0.4 would be installed (unless you used to --home
> or --prefix options when running python setup.py install).  This means
> that unless you set PYTHONPATH you will always get Leopards older
> numpy, which will cause scipy to fail on building.
>
> Note, this is very different than on Tiger where the builtin python
> did not include many third party packages.  Leopard includes quite a
> few (for better or worse).
>
> Brian
>
>
>
> On Nov 9, 2007 2:33 PM, Daniel Yarlett <yarlett at psych.stanford.edu>  
> wrote:
>> hi,
>>
>> apologies if this is not the appropriate forum for posting this,  
>> but i
>> am having some trouble building and installing scipy 0.6.0 on a PPC
>> Mac that i have just upgraded to OS X 10.5 (this version of scipy
>> previously worked fine on the same machine under OS X 10.4).
>>
>> numpy 1.0.4 seems to install fine with "python setup.py install".
>>
>> however, when i try to install scipy 0.6.0 i get the output (tail
>> only) included at the end of this email.
>>
>> here are my compiler versions:
>>
>> gcc -- version
>> gcc (GCC) 4.3.0 20071026 (experimental)
>>
>> gfortran -- version
>> GNU Fortran (GCC) 4.3.0 20071026 (experimental)
>>
>> i remember that on previous versions of OS X you could use gcc_select
>> to specify v3.4 of gcc to make the build go through. unfortunately
>> gcc_select seems to have been deprecated in Leopard so i haven't been
>> able to try that. don't know if that's a red herring or not.
>>
>> also, fwiw, i just upgraded an intel macbook prop from OS X 10.4 to
>> 10.5 and had no problems at all with the pre-existing versions of
>> numpy and scipy.
>>
>> anyway, any help would be greatly appreciated!
>>
>> best,
>>
>> dan.
>>
>> --
>>
>> END OF OUTPUT FROM "python setup.py install" for scipy 0.6.0:
>>
>> Could not locate executable f95
>> customize AbsoftFCompiler
>> Could not locate executable f90
>> Could not locate executable f77
>> customize IBMFCompiler
>> Could not locate executable xlf90
>> Could not locate executable xlf
>> customize IntelFCompiler
>> Could not locate executable ifort
>> Could not locate executable ifc
>> customize GnuFCompiler
>> Found executable /usr/local/bin/g77
>> gnu: no Fortran 90 compiler found
>> gnu: no Fortran 90 compiler found
>> customize GnuFCompiler
>> gnu: no Fortran 90 compiler found
>> gnu: no Fortran 90 compiler found
>> customize GnuFCompiler using build_clib
>> building 'superlu_src' library
>> compiling C sources
>> C compiler: gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/
>> MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp- 
>> precomp -
>> mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3
>>
>> compile options: '-DUSE_VENDOR_BLAS=1 -c'
>> gcc: scipy/linsolve/SuperLU/SRC/colamd.c
>> gcc: unrecognized option '-no-cpp-precomp'
>> cc1: error: unrecognized command line option "-arch"
>> cc1: error: unrecognized command line option "-arch"
>> cc1: error: unrecognized command line option "-Wno-long-double"
>> gcc: unrecognized option '-no-cpp-precomp'
>> cc1: error: unrecognized command line option "-arch"
>> cc1: error: unrecognized command line option "-arch"
>> cc1: error: unrecognized command line option "-Wno-long-double"
>> error: Command "gcc -arch ppc -arch i386 -isysroot /Developer/SDKs/
>> MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp- 
>> precomp -
>> mno-fused-madd -fno-common -dynamic -DNDEBUG -g -O3 -
>> DUSE_VENDOR_BLAS=1 -c scipy/linsolve/SuperLU/SRC/colamd.c -o build/
>> temp.macosx-10.3-fat-2.5/scipy/linsolve/SuperLU/SRC/colamd.o" failed
>> with exit status 1
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From j.reid at mail.cryst.bbk.ac.uk  Sat Nov 10 12:26:39 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Sat, 10 Nov 2007 17:26:39 +0000
Subject: [SciPy-user] pdf of multivariate normal
Message-ID: <fh4pm3$pvn$1@ger.gmane.org>

Am I right that this doesn't exist in scipy/numpy? Obviously it isn't 
too hard to code myself but I was surprised it wasn't there. If it is 
there, how could I have found it easily?

John.



From berthe.loic at gmail.com  Sat Nov 10 12:52:36 2007
From: berthe.loic at gmail.com (LB)
Date: Sat, 10 Nov 2007 09:52:36 -0800
Subject: [SciPy-user] pdf of multivariate normal
In-Reply-To: <fh4pm3$pvn$1@ger.gmane.org>
References: <fh4pm3$pvn$1@ger.gmane.org>
Message-ID: <1194717156.583578.152820@o3g2000hsb.googlegroups.com>

Have a look at the numpy.random module :

>>> info(numpy.random.multivariate_normal)
Return an array containing multivariate normally distributed random
numbers
with specified mean and covariance.

multivariate_normal(mean, cov) -> random values
multivariate_normal(mean, cov, [m, n, ...]) -> random values

mean must be a 1 dimensional array. cov must be a square two
dimensional
array with the same number of rows and columns as mean has elements.

The first form returns a single 1-D array containing a multivariate
normal.

The second form returns an array of shape (m, n, ..., cov.shape[0]).
In this case, output[i,j,...,:] is a 1-D array containing a
multivariate
normal.

--
LB



From j.reid at mail.cryst.bbk.ac.uk  Sat Nov 10 13:44:50 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Sat, 10 Nov 2007 18:44:50 +0000
Subject: [SciPy-user] pdf of multivariate normal
In-Reply-To: <1194717156.583578.152820@o3g2000hsb.googlegroups.com>
References: <fh4pm3$pvn$1@ger.gmane.org>
	<1194717156.583578.152820@o3g2000hsb.googlegroups.com>
Message-ID: <fh4u8o$6v7$1@ger.gmane.org>

LB wrote:
>>>> info(numpy.random.multivariate_normal)
> Return an array containing multivariate normally distributed random
> numbers

Thanks but I don't want random variates, I would like to calculate the 
probability density function.



From j.reid at mail.cryst.bbk.ac.uk  Sat Nov 10 13:55:47 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Sat, 10 Nov 2007 18:55:47 +0000
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
Message-ID: <fh4uta$8of$1@ger.gmane.org>

Could someone confirm I've got this right?

In a multivariate normal (MVN) the covariance is  E(X.X') - E(X).E(X'). 
I sample many times from a MVN with given mean and covariance. I then 
calculate the covariance of the sample. This covariance matrix should 
show no bias to be larger or smaller than the distribution's covariance 
but I see that some entries are always larger and others are smaller.

See this code:

from numpy.random import multivariate_normal
from numpy.linalg import inv
from numpy import outer

mu = [10.,10.]
precision = [[.6, .3], [.4, .3]]
sigma = inv(precision)
sample_size = 10000
num_tests = 100

for test in xrange(num_tests):
   samples = multivariate_normal(mu, sigma, [sample_size])
   sample_mean = samples.sum(axis=0)/sample_size
   #print (sample_mean - mu) > 0.
   sample_covariance = sum(outer(sample-sample_mean, sample-sample_mean) 
for sample in samples) / sample_size
   print (sample_covariance - sigma) > 0


'True' is printed when the entry is larger and False otherwise

Did I get this wrong or is this acceptable behaviour for a random 
variate generator or is it a bug?

Thanks,
John.



From matthieu.brucher at gmail.com  Sat Nov 10 14:00:54 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sat, 10 Nov 2007 20:00:54 +0100
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
In-Reply-To: <fh4uta$8of$1@ger.gmane.org>
References: <fh4uta$8of$1@ger.gmane.org>
Message-ID: <e76aa17f0711101100x23afbd6bk27ec2c2d60c5e4f7@mail.gmail.com>

IIRC, you computed the biaised variance, so it is normal to see biais. Try
dividing by (sample_size-1), it should be the unbiaised estimator.

Matthieu

2007/11/10, John Reid <j.reid at mail.cryst.bbk.ac.uk>:
>
> Could someone confirm I've got this right?
>
> In a multivariate normal (MVN) the covariance is  E(X.X') - E(X).E(X').
> I sample many times from a MVN with given mean and covariance. I then
> calculate the covariance of the sample. This covariance matrix should
> show no bias to be larger or smaller than the distribution's covariance
> but I see that some entries are always larger and others are smaller.
>
> See this code:
>
> from numpy.random import multivariate_normal
> from numpy.linalg import inv
> from numpy import outer
>
> mu = [10.,10.]
> precision = [[.6, .3], [.4, .3]]
> sigma = inv(precision)
> sample_size = 10000
> num_tests = 100
>
> for test in xrange(num_tests):
>    samples = multivariate_normal(mu, sigma, [sample_size])
>    sample_mean = samples.sum(axis=0)/sample_size
>    #print (sample_mean - mu) > 0.
>    sample_covariance = sum(outer(sample-sample_mean, sample-sample_mean)
> for sample in samples) / sample_size
>    print (sample_covariance - sigma) > 0
>
>
> 'True' is printed when the entry is larger and False otherwise
>
> Did I get this wrong or is this acceptable behaviour for a random
> variate generator or is it a bug?
>
> Thanks,
> John.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071110/ddfcd15f/attachment.html>

From j.reid at mail.cryst.bbk.ac.uk  Sat Nov 10 19:18:45 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Sun, 11 Nov 2007 00:18:45 +0000
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
In-Reply-To: <e76aa17f0711101100x23afbd6bk27ec2c2d60c5e4f7@mail.gmail.com>
References: <fh4uta$8of$1@ger.gmane.org>
	<e76aa17f0711101100x23afbd6bk27ec2c2d60c5e4f7@mail.gmail.com>
Message-ID: <fh5hqg$oat$1@ger.gmane.org>

Matthieu Brucher wrote:
> IIRC, you computed the biaised variance, so it is normal to see biais. 
> Try dividing by (sample_size-1), it should be the unbiaised estimator.
> 
That wouldn't change the sign of the answer. In any case I get the same 
results when dividing by sample_size-1.



From bryan.nollett at gmail.com  Sat Nov 10 21:26:48 2007
From: bryan.nollett at gmail.com (Bryan Nollett)
Date: Sat, 10 Nov 2007 20:26:48 -0600
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
In-Reply-To: <fh4uta$8of$1@ger.gmane.org>
References: <fh4uta$8of$1@ger.gmane.org>
Message-ID: <28fb541d0711101826y6b443a13o5c122d5263e9b643@mail.gmail.com>

The covariance matrix provided in sigma is not symmetric.  Forcing
symmetry by changing the definition of sigma to
  sigma = inv(precision)
  sigma[1,0] = sigma[0,1]
yields behavior that I suspect is in line with your expectations.

Perhaps a check for symmetry in the cov argument would be desirable
addition to multivariate_normal?

Bryan


On Nov 10, 2007 12:55 PM, John Reid <j.reid at mail.cryst.bbk.ac.uk> wrote:
> In a multivariate normal (MVN) the covariance is  E(X.X') - E(X).E(X').
> I sample many times from a MVN with given mean and covariance. I then
> calculate the covariance of the sample. This covariance matrix should
> show no bias to be larger or smaller than the distribution's covariance
> but I see that some entries are always larger and others are smaller.


From j.reid at mail.cryst.bbk.ac.uk  Sun Nov 11 05:32:44 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Sun, 11 Nov 2007 10:32:44 +0000
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
In-Reply-To: <28fb541d0711101826y6b443a13o5c122d5263e9b643@mail.gmail.com>
References: <fh4uta$8of$1@ger.gmane.org>
	<28fb541d0711101826y6b443a13o5c122d5263e9b643@mail.gmail.com>
Message-ID: <fh6lpn$rl4$1@ger.gmane.org>

Bryan Nollett wrote:
> The covariance matrix provided in sigma is not symmetric.

Good point. I needed some test covariance matrices and I completely 
forgot the symmetric requirement. There is no way to draw from the 
Wishart or Inv-Wishart distributions is there? Their pdfs would be handy 
too.

Thanks,
John.



From matthieu.brucher at gmail.com  Sun Nov 11 06:21:11 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sun, 11 Nov 2007 12:21:11 +0100
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
In-Reply-To: <fh6lpn$rl4$1@ger.gmane.org>
References: <fh4uta$8of$1@ger.gmane.org>
	<28fb541d0711101826y6b443a13o5c122d5263e9b643@mail.gmail.com>
	<fh6lpn$rl4$1@ger.gmane.org>
Message-ID: <e76aa17f0711110321n285080f6t42f4cd2c64664195@mail.gmail.com>

Did you try the rvs method of those classes ?

Matthieu

2007/11/11, John Reid <j.reid at mail.cryst.bbk.ac.uk>:
>
> Bryan Nollett wrote:
> > The covariance matrix provided in sigma is not symmetric.
>
> Good point. I needed some test covariance matrices and I completely
> forgot the symmetric requirement. There is no way to draw from the
> Wishart or Inv-Wishart distributions is there? Their pdfs would be handy
> too.
>
> Thanks,
> John.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071111/c16c8b19/attachment.html>

From j.reid at mail.cryst.bbk.ac.uk  Sun Nov 11 10:19:40 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Sun, 11 Nov 2007 15:19:40 +0000
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
In-Reply-To: <e76aa17f0711110321n285080f6t42f4cd2c64664195@mail.gmail.com>
References: <fh4uta$8of$1@ger.gmane.org>	<28fb541d0711101826y6b443a13o5c122d5263e9b643@mail.gmail.com>	<fh6lpn$rl4$1@ger.gmane.org>
	<e76aa17f0711110321n285080f6t42f4cd2c64664195@mail.gmail.com>
Message-ID: <fh76jn$5qj$1@ger.gmane.org>

Matthieu Brucher wrote:
> Did you try the rvs method of those classes ?

Of which classes? I couldn't see anything in scipy.stats or 
scipy.stats.distributions to do with the wishart distribution.



From vel.accel at gmail.com  Sun Nov 11 17:10:48 2007
From: vel.accel at gmail.com (dieter h.)
Date: Sun, 11 Nov 2007 17:10:48 -0500
Subject: [SciPy-user] Sandbox TimeSeries: request
Message-ID: <1e52e0880711111410g150abf7dhe30b5cfc3adb0fbc@mail.gmail.com>

Hi Pierre and Matt

If I personally had the time to dig in to the module (and extension
writing itself), I'd be incorporating this request myself.

Functionality that would convert an array of epochs to
datetime.datetime|mx.datetime|matplotlib.Dates objects  would be a big
plus. I suggest it in case you might also find it worthy. In the mean
time I'll get by. :-)

Thank you
-Dieter


From pgmdevlist at gmail.com  Sun Nov 11 17:40:08 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Sun, 11 Nov 2007 17:40:08 -0500
Subject: [SciPy-user] Sandbox TimeSeries: request
In-Reply-To: <1e52e0880711111410g150abf7dhe30b5cfc3adb0fbc@mail.gmail.com>
References: <1e52e0880711111410g150abf7dhe30b5cfc3adb0fbc@mail.gmail.com>
Message-ID: <200711111740.08260.pgmdevlist@gmail.com>


> Functionality that would convert an array of epochs to
> datetime.datetime|mx.datetime|matplotlib.Dates objects  would be a big
> plus.

You can already transform a single Date object to datetime with datetime. It 
would be pretty straightforward to add a todatetime() method to the DateArray 
class, which would solve the first request. 

For matplotlib, I'd suggest you give the plotlib module a try: we have 
implemented a dynamic location/formatting of ticks that can be pretty 
useful...



From rshepard at appl-ecosys.com  Sun Nov 11 18:52:14 2007
From: rshepard at appl-ecosys.com (Rich Shepard)
Date: Sun, 11 Nov 2007 15:52:14 -0800 (PST)
Subject: [SciPy-user] Problem building SciPy-0.6.0
Message-ID: <Pine.LNX.4.64.0711111547510.24388@salmo.appl-ecosys.com>

   I upgraded my notebook from Slackware-11.0 to -12.0 so libc, gcc, etc. are
different. I've built ATLAS (and copied the *.a and *.so libraries to
/usr/local/lib), then sucessfully built and installed NumPy-1.0.4. But, when
I try 'python setup.py install' from the scipy-0.6.0 directory the attempt
immediately fails:

Traceback (most recent call last):
   File "setup.py", line 53, in <module>
     setup_package()
   File "setup.py", line 28, in setup_package
     from numpy.distutils.core import setup
   File "/usr/lib/python2.5/site-packages/numpy/__init__.py", line 43, in <module>
     import linalg
   File "/usr/lib/python2.5/site-packages/numpy/linalg/__init__.py", line 4, in <module>
     from linalg import *
   File "/usr/lib/python2.5/site-packages/numpy/linalg/linalg.py", line 25, in <module>
     from numpy.linalg import lapack_lite
ImportError: liblapack.so: cannot open shared object file: No such file or directory

   liblapack.so is in both /usr/local/lib and
~/python/ATLAS/ATLAS-linux/lib/.

   What am I missing here?

Rich

-- 
Richard B. Shepard, Ph.D.               |  Integrity            Credibility
Applied Ecosystem Services, Inc.        |            Innovation
<http://www.appl-ecosys.com>     Voice: 503-667-4517      Fax: 503-667-8863


From rshepard at appl-ecosys.com  Sun Nov 11 19:29:58 2007
From: rshepard at appl-ecosys.com (Rich Shepard)
Date: Sun, 11 Nov 2007 16:29:58 -0800 (PST)
Subject: [SciPy-user] Problem building SciPy-0.6.0 -- FIXED
In-Reply-To: <Pine.LNX.4.64.0711111547510.24388@salmo.appl-ecosys.com>
References: <Pine.LNX.4.64.0711111547510.24388@salmo.appl-ecosys.com>
Message-ID: <Pine.LNX.4.64.0711111629040.24388@salmo.appl-ecosys.com>

On Sun, 11 Nov 2007, Rich Shepard wrote:

>   What am I missing here?

   Found it: python was looking in /usr/lib, and not in /usr/local/lib.
Softlinks fixed this problem.

Rich

-- 
Richard B. Shepard, Ph.D.               |  Integrity            Credibility
Applied Ecosystem Services, Inc.        |            Innovation
<http://www.appl-ecosys.com>     Voice: 503-667-4517      Fax: 503-667-8863


From david at ar.media.kyoto-u.ac.jp  Sun Nov 11 22:03:02 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 12 Nov 2007 12:03:02 +0900
Subject: [SciPy-user] Bias in numpy.random.multivariate_normal?
In-Reply-To: <fh76jn$5qj$1@ger.gmane.org>
References: <fh4uta$8of$1@ger.gmane.org>	<28fb541d0711101826y6b443a13o5c122d5263e9b643@mail.gmail.com>	<fh6lpn$rl4$1@ger.gmane.org>	<e76aa17f0711110321n285080f6t42f4cd2c64664195@mail.gmail.com>
	<fh76jn$5qj$1@ger.gmane.org>
Message-ID: <4737C266.1080905@ar.media.kyoto-u.ac.jp>

John Reid wrote:
> Matthieu Brucher wrote:
>   
>> Did you try the rvs method of those classes ?
>>     
>
> Of which classes? I couldn't see anything in scipy.stats or 
> scipy.stats.distributions to do with the wishart distribution.
>   
I don't think there is. This would be handy, indeed. So handy that I 
have just created a ticket for it :)

http://scipy.org/scipy/scipy/ticket/536

cheers,

David


From c.j.lee at tnw.utwente.nl  Mon Nov 12 06:14:50 2007
From: c.j.lee at tnw.utwente.nl (Chris Lee)
Date: Mon, 12 Nov 2007 12:14:50 +0100
Subject: [SciPy-user] instrumentation control question
Message-ID: <473835AA.20801@tnw.utwente.nl>

Hi, I know this is a bit off-topic but I wondered if anyone here had any 
experience with the following problem.

Part of our work involves controlling a bunch of lab instrumentation, 
which so far has worked beautifully using python + scipy/numpy + pyVISA. 
Then we got a stepper motor which does not comes with an activeX control 
and (as far as I can tell) nothing else.

This means using the com extensions for win32 and trying to communicate 
with the activeX control. All is not progressing well (we can find the 
control but send no data to it, and it certainly won't talk back to us). 
Does anyone have any experience with this? If so, would you mind sending 
me some code snippets for gaining control of the activeX element?

Thanks very much and my appologies for the not-so-on-topic post (I 
actually think instrumentation control would be a great addition to 
scipy and if I become competent may look at contributing code in that 
direction).

Cheers
Chris


From sandricionut at yahoo.com  Mon Nov 12 09:07:57 2007
From: sandricionut at yahoo.com (sandric ionut)
Date: Mon, 12 Nov 2007 06:07:57 -0800 (PST)
Subject: [SciPy-user] random gaussian array
Message-ID: <834845.12895.qm@web51304.mail.re2.yahoo.com>

Hi:
 
I want to create an random Gaussian 2D array with an exact size. For example I want my array to look like:
 
2 3 4 0 1 1
2 3 3 4 1 1
3 2 5 6 7 8

all the values should be randomly generated from a Gaussian distribution with mean and standard deviation
 
Is it possible to do it with scipy or numpy?
 
Thank you
 
Ionut

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071112/4909bd56/attachment.html>

From matthieu.brucher at gmail.com  Mon Nov 12 09:11:08 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Mon, 12 Nov 2007 15:11:08 +0100
Subject: [SciPy-user] random gaussian array
In-Reply-To: <834845.12895.qm@web51304.mail.re2.yahoo.com>
References: <834845.12895.qm@web51304.mail.re2.yahoo.com>
Message-ID: <e76aa17f0711120611u6d4d3dcy9764c2a0f48e067@mail.gmail.com>

Yes, look into numpy.random -> norm

Matthieu

2007/11/12, sandric ionut <sandricionut at yahoo.com>:
>
> Hi:
>
> I want to create an random Gaussian 2D array with an exact size. For
> example I want my array to look like:
>
> 2 3 4 0 1 1
> 2 3 3 4 1 1
> 3 2 5 6 7 8
>
> all the values should be randomly generated from a Gaussian distribution
> with mean and standard deviation
>
> Is it possible to do it with scipy or numpy?
>
> Thank you
>
> Ionut
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071112/4a15de58/attachment.html>

From berthe.loic at gmail.com  Mon Nov 12 11:52:35 2007
From: berthe.loic at gmail.com (LB)
Date: Mon, 12 Nov 2007 16:52:35 -0000
Subject: [SciPy-user] nullclines for nonlinear ODEs
In-Reply-To: <1193769594.019129.155320@d55g2000hsg.googlegroups.com>
References: <88e473830710300926l66c350c3k72b7fda9e6fd5c31@mail.gmail.com>
	<88e473830710300928r68f376b8k2a6a5c76b04c11c1@mail.gmail.com>
	<1193769594.019129.155320@d55g2000hsg.googlegroups.com>
Message-ID: <1194886355.028155.281910@v3g2000hsg.googlegroups.com>

Just for information, I've put a first version of this tutorial on the
Wiki : http://www.scipy.org/LoktaVolterraTutorial.

I don't talk about nullclines for the moment but feel to complete it -
or to correct it if you find some mistakes.

--
LB



From jeremit0 at gmail.com  Mon Nov 12 12:06:31 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Mon, 12 Nov 2007 12:06:31 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
Message-ID: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>

Hello all.  I recently upgraded scipy from the svn distribution and
tried installing it.  I followed the instructions on
http://scipy.org/Installing_SciPy/Mac_OS_X

and downloaded the gfortran binaries from the att.com site.  When I run

python setup.py build

it compiles for awhile but fails.  The last few messages are copied at
the end of this message.  Is another compiler (i.e. f95) needed?  Do I
just need to rename the gfortran executable?  I tried making a symlink
 g95->gfortran but it didn't work.  Does anyone have any suggestions?

Thanks,
Jeremy

$ gcc --version
i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5465)
Copyright (C) 2005 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

$ gfortran --version
GNU Fortran (GCC) 4.2.1
Copyright (C) 2007 Free Software Foundation, Inc.

GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
You may redistribute copies of GNU Fortran
under the terms of the GNU General Public License.
For more information about these matters, see the file named COPYING

$ python setup.py build
?
compiling Fortran sources
Fortran f77 compiler: f95 -fixed -O4 -target=native
Fortran f90 compiler: f95 -O4 -target=native
Fortran fix compiler: f95 -fixed -O4 -target=native
creating build/temp.macosx-10.5-i386-2.5
creating build/temp.macosx-10.5-i386-2.5/scipy
creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack
creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack
compile options: '-c'
f95:f77: scipy/fftpack/dfftpack/dcosqb.f
sh: f95: command not found
sh: f95: command not found
error: Command "f95 -fixed -O4 -target=native -c -c
scipy/fftpack/dfftpack/dcosqb.f -o
build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack/dcosqb.o"
failed with exit status 127


From hasslerjc at comcast.net  Mon Nov 12 13:00:32 2007
From: hasslerjc at comcast.net (John Hassler)
Date: Mon, 12 Nov 2007 13:00:32 -0500
Subject: [SciPy-user] instrumentation control question
In-Reply-To: <473835AA.20801@tnw.utwente.nl>
References: <473835AA.20801@tnw.utwente.nl>
Message-ID: <473894C0.7000005@comcast.net>

I've used ctypes to access .dll libraries, but I've never used activex.  
Andrew Straw has written a very nice library for Measurement Computing 
hardware.  Perhaps something there would be useful:
http://code.astraw.com/

Do you have a hardware stepper driver, or are you trying to do it all in 
software?
john

Chris Lee wrote:
> Hi, I know this is a bit off-topic but I wondered if anyone here had any 
> experience with the following problem.
>
> Part of our work involves controlling a bunch of lab instrumentation, 
> which so far has worked beautifully using python + scipy/numpy + pyVISA. 
> Then we got a stepper motor which does not comes with an activeX control 
> and (as far as I can tell) nothing else.
>
> This means using the com extensions for win32 and trying to communicate 
> with the activeX control. All is not progressing well (we can find the 
> control but send no data to it, and it certainly won't talk back to us). 
> Does anyone have any experience with this? If so, would you mind sending 
> me some code snippets for gaining control of the activeX element?
>
> Thanks very much and my appologies for the not-so-on-topic post (I 
> actually think instrumentation control would be a great addition to 
> scipy and if I become competent may look at contributing code in that 
> direction).
>
> Cheers
> Chris
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>   


From dominique.orban at gmail.com  Mon Nov 12 13:19:31 2007
From: dominique.orban at gmail.com (Dominique Orban)
Date: Mon, 12 Nov 2007 13:19:31 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
Message-ID: <8793ae6e0711121019k35291310q664910303c6cdac7@mail.gmail.com>

On 11/12/07, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> Hello all.  I recently upgraded scipy from the svn distribution and
> tried installing it.  I followed the instructions on
> http://scipy.org/Installing_SciPy/Mac_OS_X
>
> and downloaded the gfortran binaries from the att.com site.  When I run
>
> python setup.py build
>
> it compiles for awhile but fails.  The last few messages are copied at
> the end of this message.  Is another compiler (i.e. f95) needed?  Do I
> just need to rename the gfortran executable?  I tried making a symlink
>  g95->gfortran but it didn't work.  Does anyone have any suggestions?
>
> Thanks,
> Jeremy
>
> $ gcc --version
> i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5465)
> Copyright (C) 2005 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions.  There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>
> $ gfortran --version
> GNU Fortran (GCC) 4.2.1
> Copyright (C) 2007 Free Software Foundation, Inc.
>
> GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
> You may redistribute copies of GNU Fortran
> under the terms of the GNU General Public License.
> For more information about these matters, see the file named COPYING
>
> $ python setup.py build
> ?
> compiling Fortran sources
> Fortran f77 compiler: f95 -fixed -O4 -target=native
> Fortran f90 compiler: f95 -O4 -target=native
> Fortran fix compiler: f95 -fixed -O4 -target=native
> creating build/temp.macosx-10.5-i386-2.5
> creating build/temp.macosx-10.5-i386-2.5/scipy
> creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack
> creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack
> compile options: '-c'
> f95:f77: scipy/fftpack/dfftpack/dcosqb.f
> sh: f95: command not found
> sh: f95: command not found
> error: Command "f95 -fixed -O4 -target=native -c -c
> scipy/fftpack/dfftpack/dcosqb.f -o
> build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack/dcosqb.o"
> failed with exit status 127
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>

Hi Jeremy,

did you perhaps forget to specify that you want to use gfortran to
build SciPy? You can do it as so:

  python setup.py config_fc --fcompiler=gfortran build

See INSTALL.txt for more information.

Good luck,
Dominique

From C.J.Lee at tnw.utwente.nl  Mon Nov 12 13:26:06 2007
From: C.J.Lee at tnw.utwente.nl (C.J.Lee at tnw.utwente.nl)
Date: Mon, 12 Nov 2007 19:26:06 +0100
Subject: [SciPy-user] instrumentation control question
References: <473835AA.20801@tnw.utwente.nl> <473894C0.7000005@comcast.net>
Message-ID: <9E7F7554631C0047AB598527A4E9569B6F8121@tnx4.dynamic.tnw.utwente.nl>

Well actually it is a DC motor that is controlled via a USB "block" that can be used to drive either stepper motors, DC servos, or piezos, or a combination (depending on the number of channels)

I have discovered where the dlls are hidden but have not managed to drive it using ctypes (I have had success with this using a different motor control/motor combination)

I will take a look at andrew's web site.

Cheers
Chris

PS John, sorry for the double email but I didn't reply to the list on the first occasion


-----Original Message-----
From: John Hassler [mailto:hasslerjc at comcast.net]
Sent: Mon 11/12/2007 7:00 PM
To: Lee, Chris (TNW); SciPy Users List
Subject: Re: [SciPy-user] instrumentation control question
 
I've used ctypes to access .dll libraries, but I've never used activex.  
Andrew Straw has written a very nice library for Measurement Computing 
hardware.  Perhaps something there would be useful:
http://code.astraw.com/

Do you have a hardware stepper driver, or are you trying to do it all in 
software?
john

Chris Lee wrote:
> Hi, I know this is a bit off-topic but I wondered if anyone here had any 
> experience with the following problem.
>
> Part of our work involves controlling a bunch of lab instrumentation, 
> which so far has worked beautifully using python + scipy/numpy + pyVISA. 
> Then we got a stepper motor which does not comes with an activeX control 
> and (as far as I can tell) nothing else.
>
> This means using the com extensions for win32 and trying to communicate 
> with the activeX control. All is not progressing well (we can find the 
> control but send no data to it, and it certainly won't talk back to us). 
> Does anyone have any experience with this? If so, would you mind sending 
> me some code snippets for gaining control of the activeX element?
>
> Thanks very much and my appologies for the not-so-on-topic post (I 
> actually think instrumentation control would be a great addition to 
> scipy and if I become competent may look at contributing code in that 
> direction).
>
> Cheers
> Chris
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>   



From ellisonbg.net at gmail.com  Mon Nov 12 13:39:04 2007
From: ellisonbg.net at gmail.com (Brian Granger)
Date: Mon, 12 Nov 2007 11:39:04 -0700
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
Message-ID: <6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>

Are you using the builtin python?
Are you using the numpy that comes with the builtin python?

If so, you will need to upgrade numpy first.  But, there are some
pitfalls along this path as well.  See recent threads on the numpy
list and the pythonmac-sig list for details.  Or, just use the python
binary from python.org and install numpy/scipy from scratch.

Brian





On Nov 12, 2007 10:06 AM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> Hello all.  I recently upgraded scipy from the svn distribution and
> tried installing it.  I followed the instructions on
> http://scipy.org/Installing_SciPy/Mac_OS_X
>
> and downloaded the gfortran binaries from the att.com site.  When I run
>
> python setup.py build
>
> it compiles for awhile but fails.  The last few messages are copied at
> the end of this message.  Is another compiler (i.e. f95) needed?  Do I
> just need to rename the gfortran executable?  I tried making a symlink
>  g95->gfortran but it didn't work.  Does anyone have any suggestions?
>
> Thanks,
> Jeremy
>
> $ gcc --version
> i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5465)
> Copyright (C) 2005 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions.  There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>
> $ gfortran --version
> GNU Fortran (GCC) 4.2.1
> Copyright (C) 2007 Free Software Foundation, Inc.
>
> GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
> You may redistribute copies of GNU Fortran
> under the terms of the GNU General Public License.
> For more information about these matters, see the file named COPYING
>
> $ python setup.py build
> ?
> compiling Fortran sources
> Fortran f77 compiler: f95 -fixed -O4 -target=native
> Fortran f90 compiler: f95 -O4 -target=native
> Fortran fix compiler: f95 -fixed -O4 -target=native
> creating build/temp.macosx-10.5-i386-2.5
> creating build/temp.macosx-10.5-i386-2.5/scipy
> creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack
> creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack
> compile options: '-c'
> f95:f77: scipy/fftpack/dfftpack/dcosqb.f
> sh: f95: command not found
> sh: f95: command not found
> error: Command "f95 -fixed -O4 -target=native -c -c
> scipy/fftpack/dfftpack/dcosqb.f -o
> build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack/dcosqb.o"
> failed with exit status 127
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From jeremit0 at gmail.com  Mon Nov 12 14:23:03 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Mon, 12 Nov 2007 14:23:03 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <8793ae6e0711121019k35291310q664910303c6cdac7@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<8793ae6e0711121019k35291310q664910303c6cdac7@mail.gmail.com>
Message-ID: <3db594f70711121123l1919a490q1ad159e0ea8a4f0c@mail.gmail.com>

On Nov 12, 2007 1:19 PM, Dominique Orban <dominique.orban at gmail.com> wrote:
>
> On 11/12/07, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> > Hello all.  I recently upgraded scipy from the svn distribution and
> > tried installing it.  I followed the instructions on
> > http://scipy.org/Installing_SciPy/Mac_OS_X
> >
> > and downloaded the gfortran binaries from the att.com site.  When I run
> >
> > python setup.py build
> >
> > it compiles for awhile but fails.  The last few messages are copied at
> > the end of this message.  Is another compiler (i.e. f95) needed?  Do I
> > just need to rename the gfortran executable?  I tried making a symlink
> >  g95->gfortran but it didn't work.  Does anyone have any suggestions?
> >
> > Thanks,
> > Jeremy
> >
> > $ gcc --version
> > i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5465)
> > Copyright (C) 2005 Free Software Foundation, Inc.
> > This is free software; see the source for copying conditions.  There is NO
> > warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> >
> > $ gfortran --version
> > GNU Fortran (GCC) 4.2.1
> > Copyright (C) 2007 Free Software Foundation, Inc.
> >
> > GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
> > You may redistribute copies of GNU Fortran
> > under the terms of the GNU General Public License.
> > For more information about these matters, see the file named COPYING
> >
> > $ python setup.py build
> > ?
> > compiling Fortran sources
> > Fortran f77 compiler: f95 -fixed -O4 -target=native
> > Fortran f90 compiler: f95 -O4 -target=native
> > Fortran fix compiler: f95 -fixed -O4 -target=native
> > creating build/temp.macosx-10.5-i386-2.5
> > creating build/temp.macosx-10.5-i386-2.5/scipy
> > creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack
> > creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack
> > compile options: '-c'
> > f95:f77: scipy/fftpack/dfftpack/dcosqb.f
> > sh: f95: command not found
> > sh: f95: command not found
> > error: Command "f95 -fixed -O4 -target=native -c -c
> > scipy/fftpack/dfftpack/dcosqb.f -o
> > build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack/dcosqb.o"
> > failed with exit status 127
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
>
> Hi Jeremy,
>
> did you perhaps forget to specify that you want to use gfortran to
> build SciPy? You can do it as so:
>
>   python setup.py config_fc --fcompiler=gfortran build
>
> See INSTALL.txt for more information.

Duh!  I did forget that important detail.  I'll look more closely at
the instructions on the wiki next time :P.  I'll post again if there
are troubles.

Thanks,
Jeremy


From jeremit0 at gmail.com  Mon Nov 12 14:25:24 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Mon, 12 Nov 2007 14:25:24 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
Message-ID: <3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>

On Nov 12, 2007 1:39 PM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> Are you using the builtin python?
> Are you using the numpy that comes with the builtin python?
>
> If so, you will need to upgrade numpy first.  But, there are some
> pitfalls along this path as well.  See recent threads on the numpy
> list and the pythonmac-sig list for details.  Or, just use the python
> binary from python.org and install numpy/scipy from scratch.
>
> Brian

I am using the builtin python.  I did build numpy myself from svn, but
it looks like python is using builtin version.  How can I make it use
my compiled version instead?

I will take a look at the pythonmac-sig email list and see what is
known about scipy on Leopard.

Thanks again,
Jeremy


From mhearne at usgs.gov  Mon Nov 12 15:26:02 2007
From: mhearne at usgs.gov (Michael Hearne)
Date: Mon, 12 Nov 2007 13:26:02 -0700
Subject: [SciPy-user] Bundling numpy/scipy with applications
Message-ID: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>

I'm creating a Python application that is internal to my  
organization, but will be installed on both Linux and Mac OS X  
machines.  The application depends heavily on a number of "non-pure"  
modules (those that have C/C++/FORTRAN components), like numpy,  
scipy, gdal, etc.

What is the most "pythonic" way to bundle these types of modules  
inside my application?

I've investigated dist_utils and setuptools, and I don't see an easy  
way with those tools to include build instructions for packages built  
with autconf tools.

Is my only recourse to write a custom install script that calls the  
"configure;make;make install" from a shell?

--Mike



------------------------------------------------------
Michael Hearne
mhearne at usgs.gov
(303) 273-8620
USGS National Earthquake Information Center
1711 Illinois St. Golden CO 80401
Senior Software Engineer
Synergetics, Inc.
------------------------------------------------------


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071112/fbf8dbdb/attachment.html>

From lists.steve at arachnedesign.net  Mon Nov 12 16:02:48 2007
From: lists.steve at arachnedesign.net (Steve Lianoglou)
Date: Mon, 12 Nov 2007 16:02:48 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
Message-ID: <4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>

> I am using the builtin python.  I did build numpy myself from svn, but
> it looks like python is using builtin version.  How can I make it use
> my compiled version instead?
>
> I will take a look at the pythonmac-sig email list and see what is
> known about scipy on Leopard.

If I remember correctly, you have to setup your PYTHONPATH envi var to  
include the libraries in /Library/Python/2.5/site-packages before the  
ones that are installed into the system path (which I think are buried  
somewhere in /System/Library/Frameworks/Python.framework/...).

(The normal `python setup.py install` command will drop then into  / 
Library/Python/2.5/site-packages)

Also, FWIW, I've had better luck using the gfortran binary from:
http://r.research.att.com/tools

Over the one that I think is suggested at the tutorial (from  
hpc.sf.net) even though they haven't been upgraded for Leopard yet  
(which I don't think they even need to be).

HTH,
-steve


From aisaac at american.edu  Mon Nov 12 16:46:42 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Mon, 12 Nov 2007 16:46:42 -0500
Subject: [SciPy-user] nullclines for nonlinear ODEs
In-Reply-To: <1194886355.028155.281910@v3g2000hsg.googlegroups.com>
References: <88e473830710300926l66c350c3k72b7fda9e6fd5c31@mail.gmail.com><88e473830710300928r68f376b8k2a6a5c76b04c11c1@mail.gmail.com><1193769594.019129.155320@d55g2000hsg.googlegroups.com><1194886355.028155.281910@v3g2000hsg.googlegroups.com>
Message-ID: <Mahogany-0.67.0-804-20071112-164642.00@american.edu>

On Mon, 12 Nov 2007, LB apparently wrote:
> http://www.scipy.org/LoktaVolterraTutorial 

After a quick look, I believe this will be very useful
for my students.

Thank you!
Alan Isaac






From jeremit0 at gmail.com  Mon Nov 12 16:51:38 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Mon, 12 Nov 2007 16:51:38 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
	<4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
Message-ID: <3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>

On Nov 12, 2007 4:02 PM, Steve Lianoglou <lists.steve at arachnedesign.net> wrote:
> > I am using the builtin python.  I did build numpy myself from svn, but
> > it looks like python is using builtin version.  How can I make it use
> > my compiled version instead?
> >
> > I will take a look at the pythonmac-sig email list and see what is
> > known about scipy on Leopard.
>
> If I remember correctly, you have to setup your PYTHONPATH envi var to
> include the libraries in /Library/Python/2.5/site-packages before the
> ones that are installed into the system path (which I think are buried
> somewhere in /System/Library/Frameworks/Python.framework/...).
>
> (The normal `python setup.py install` command will drop then into  /
> Library/Python/2.5/site-packages)

I got numpy and scipy to install and they both passed all the tests.
This is how I did it.

I followed the instructions on scipy.org for installing on Mac OS X.
I used the fortran compiler from att.com site.  I set my PYTHONPATH
as:

export PYTHONPATH="/Library/Python/2.5/site-packages:$PYTHONPATH"

Everything compiled without any problems.  Thanks for everyone's
suggestions.  I hope my results will help others.

Jeremy


From cohen at slac.stanford.edu  Mon Nov 12 17:02:24 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Mon, 12 Nov 2007 14:02:24 -0800
Subject: [SciPy-user] problem building scipy : liblapack.so not found....
Message-ID: <4738CD70.7080401@slac.stanford.edu>

I have the current SVN trunk, and I built lapack and ATLAS following the
doc in the scipy web site. I also built numpy from SVN.
Now when trying toinstall scipy I get :
[cohen at localhost scipy-svn]$ su -c 'python setup.py install'
Password:
Traceback (most recent call last):
  File "setup.py", line 92, in <module>
    setup_package()
  File "setup.py", line 63, in setup_package
    from numpy.distutils.core import setup
  File "/usr/lib/python2.5/site-packages/numpy/__init__.py", line 43, in
<module>
    import linalg
  File "/usr/lib/python2.5/site-packages/numpy/linalg/__init__.py", line
4, in <module>
    from linalg import *
  File "/usr/lib/python2.5/site-packages/numpy/linalg/linalg.py", line
25, in <module>
    from numpy.linalg import lapack_lite
ImportError: liblapack.so: cannot open shared object file: No such file
or directory

but I do have this library:
[cohen at localhost scipy-svn]$ ls -l /usr/local/lib/liblapack.so
lrwxrwxrwx 1 root root 33 2007-11-12 09:29 /usr/local/lib/liblapack.so
-> /usr/local/atlas/lib/liblapack.so
and it should be in my paths.
Moreover, I can issue the offending line without problem:
[cohen at localhost scipy-svn]$ ipython
Python 2.5 (r25:51908, Oct 19 2007, 09:47:40)
Type "copyright", "credits" or "license" for more information.

IPython 0.8.2.svn.r2848 -- An enhanced Interactive Python.
?         -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help      -> Python's own help system.
object?   -> Details about 'object'. ?object also works, ?? prints more.

In [1]: from numpy.linalg import lapack_lite

In [2]: dir(lapack_lite)
Out[2]:
['LapackError',
 '__doc__',
 '__file__',
 '__name__',
 'dgeev',
 'dgelsd',
 'dgeqrf',
 'dgesdd',
 'dgesv',
 'dgetrf',
 'dorgqr',
 'dpotrf',
 'dsyevd',
 'zgeev',
 'zgelsd',
 'zgeqrf',
 'zgesdd',
 'zgesv',
 'zgetrf',
 'zheevd',
 'zpotrf',
 'zungqr']

\What is going on?
thanks,
Johann



From david at ar.media.kyoto-u.ac.jp  Mon Nov 12 23:48:18 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 13 Nov 2007 13:48:18 +0900
Subject: [SciPy-user] Bundling numpy/scipy with applications
In-Reply-To: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>
References: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>
Message-ID: <47392C92.8020208@ar.media.kyoto-u.ac.jp>

Michael Hearne wrote:
> I'm creating a Python application that is internal to my organization, 
> but will be installed on both Linux and Mac OS X machines.  The 
> application depends heavily on a number of "non-pure" modules (those 
> that have C/C++/FORTRAN components), like numpy, scipy, gdal, etc.
>
> What is the most "pythonic" way to bundle these types of modules 
> inside my application?
>
> I've investigated dist_utils and setuptools, and I don't see an easy 
> way with those tools to include build instructions for packages built 
> with autconf tools.
>
> Is my only recourse to write a custom install script that calls the 
> "configure;make;make install" from a shell?
>
You have two problems: building and packaging. For packaging, autoconf 
will not help you much outside providing a coherence between packages, I 
think; distutils/setuptools have packaging tools, but I don't know how 
good they are.

On Linux, the packaging system depends on the distribution: since it is 
only internal, the problem is severely simplified, since you are likely 
to target only one format (rpm, deb, etc...). On Mac OS X, you have also 
tools to package .pkg and .mpkg (which are a set of .pkg), available 
freely on apple dev website; I have not used them much, but they look 
simple and easy to use for simple command line packages.

You could hack something in distutils to call configure and so on, but I 
don't think it will be pleasant. Building packages using 
distutils/setuptools is easy, and do not need much configuration. So I 
think it is more natural to use a top level tool calling 
distutils/setuptools than to use distutils as the top level tool. As a 
pythonic way, you may consider build tools, such as scons or waf, both 
written in python.

But frankly, I would not bother too much about pythonic way: since some 
tools will be shell based anyway (autotools, packaging tools), and you 
don't have to care about windows, using make and shell may actually be 
easier.

For the build part, you may take a look at my garnumpy package: it is 
essentially a set of rules for Gnu Makefiles, and it can build numpy + 
scipy, with a configurable set of dependencies (ATLAS, NETLIB 
BLAS/LAPACK, fftw, etc....). It can build both distutils and 
autotools-based packages.

http://www.ar.media.kyoto-u.ac.jp/members/david/archives/garnumpy/garnumpy-0.4.tbz2

I used it sucessfully on linux and cygwin, so I would say it should work 
on Mac OS X without too much trouble. The only thing which will be 
likely to be a pain is fat binaries (Universal). I use it to build a 
totally self contained numpy/scipy installation, which is a first step 
toward packaging. If you think it can be useful to you, don't hesitate 
to ask questions; there is also a bzr archive if you want to have access 
to the dev history of the tool

cheers,

David


From ellisonbg.net at gmail.com  Mon Nov 12 23:59:08 2007
From: ellisonbg.net at gmail.com (Brian Granger)
Date: Mon, 12 Nov 2007 21:59:08 -0700
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
	<4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
	<3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>
Message-ID: <6ce0ac130711122059x610214bey5698dde00ee9cd8c@mail.gmail.com>

Great, now you are ready for the other problem on Leopard:

sudo does not preserve the PYTHONPATH environment variable.

Thus anytime you do::

sudo python setup.py install

to install something else, your version of numpy won't be used!  This
has been reported to apple and the problem can be fixed by adding
PYTHONPATH to the appropriate section in /etc/sudoers.

Another option (see the recent thread on pythonmax-sig) is to create a
.pth file that dynamically reorders the sys.path.  This approach is
nice as it doesn't require setting PYTHONPATH.

You will run into this when installing matplotlib and pytables...if
you go that way.

Brian

On Nov 12, 2007 2:51 PM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> On Nov 12, 2007 4:02 PM, Steve Lianoglou <lists.steve at arachnedesign.net> wrote:
> > > I am using the builtin python.  I did build numpy myself from svn, but
> > > it looks like python is using builtin version.  How can I make it use
> > > my compiled version instead?
> > >
> > > I will take a look at the pythonmac-sig email list and see what is
> > > known about scipy on Leopard.
> >
> > If I remember correctly, you have to setup your PYTHONPATH envi var to
> > include the libraries in /Library/Python/2.5/site-packages before the
> > ones that are installed into the system path (which I think are buried
> > somewhere in /System/Library/Frameworks/Python.framework/...).
> >
> > (The normal `python setup.py install` command will drop then into  /
> > Library/Python/2.5/site-packages)
>
> I got numpy and scipy to install and they both passed all the tests.
> This is how I did it.
>
> I followed the instructions on scipy.org for installing on Mac OS X.
> I used the fortran compiler from att.com site.  I set my PYTHONPATH
> as:
>
> export PYTHONPATH="/Library/Python/2.5/site-packages:$PYTHONPATH"
>
> Everything compiled without any problems.  Thanks for everyone's
> suggestions.  I hope my results will help others.
>
> Jeremy
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From cohen at slac.stanford.edu  Tue Nov 13 00:58:47 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Mon, 12 Nov 2007 21:58:47 -0800
Subject: [SciPy-user] Bundling numpy/scipy with applications
In-Reply-To: <47392C92.8020208@ar.media.kyoto-u.ac.jp>
References: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>
	<47392C92.8020208@ar.media.kyoto-u.ac.jp>
Message-ID: <47393D17.9090000@slac.stanford.edu>

David Cournapeau wrote:
> Michael Hearne wrote:
>   
>> I'm creating a Python application that is internal to my organization, 
>> but will be installed on both Linux and Mac OS X machines.  The 
>> application depends heavily on a number of "non-pure" modules (those 
>> that have C/C++/FORTRAN components), like numpy, scipy, gdal, etc.
>>
>> What is the most "pythonic" way to bundle these types of modules 
>> inside my application?
>>
>> I've investigated dist_utils and setuptools, and I don't see an easy 
>> way with those tools to include build instructions for packages built 
>> with autconf tools.
>>
>> Is my only recourse to write a custom install script that calls the 
>> "configure;make;make install" from a shell?
>>
>>     
> You have two problems: building and packaging. For packaging, autoconf 
> will not help you much outside providing a coherence between packages, I 
> think; distutils/setuptools have packaging tools, but I don't know how 
> good they are.
>
> On Linux, the packaging system depends on the distribution: since it is 
> only internal, the problem is severely simplified, since you are likely 
> to target only one format (rpm, deb, etc...). On Mac OS X, you have also 
> tools to package .pkg and .mpkg (which are a set of .pkg), available 
> freely on apple dev website; I have not used them much, but they look 
> simple and easy to use for simple command line packages.
>
> You could hack something in distutils to call configure and so on, but I 
> don't think it will be pleasant. Building packages using 
> distutils/setuptools is easy, and do not need much configuration. So I 
> think it is more natural to use a top level tool calling 
> distutils/setuptools than to use distutils as the top level tool. As a 
> pythonic way, you may consider build tools, such as scons or waf, both 
> written in python.
>
> But frankly, I would not bother too much about pythonic way: since some 
> tools will be shell based anyway (autotools, packaging tools), and you 
> don't have to care about windows, using make and shell may actually be 
> easier.
>
> For the build part, you may take a look at my garnumpy package: it is 
> essentially a set of rules for Gnu Makefiles, and it can build numpy + 
> scipy, with a configurable set of dependencies (ATLAS, NETLIB 
> BLAS/LAPACK, fftw, etc....). It can build both distutils and 
> autotools-based packages.
>
> http://www.ar.media.kyoto-u.ac.jp/members/david/archives/garnumpy/garnumpy-0.4.tbz2
>
> I used it sucessfully on linux and cygwin, so I would say it should work 
> on Mac OS X without too much trouble. The only thing which will be 
> likely to be a pain is fat binaries (Universal). I use it to build a 
> totally self contained numpy/scipy installation, which is a first step 
> toward packaging. If you think it can be useful to you, don't hesitate 
> to ask questions; there is also a bzr archive if you want to have access 
> to the dev history of the tool
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   
Hi David,
I am having a hell of a time with lapack/atlas. I already posted the 
issue I currently have with scipy after following the building doc in 
scipy web site (see 
http://projects.scipy.org/pipermail/scipy-user/2007-November/014506.html). 
Besides, octave build seems to also have serious issues with my 
lapack/atlas build.

So I immediately downloaded youor package, but I seem to have a problem 
with the numpy download :
make[3]: Leaving directory 
`/data1/sources/python/garnumpy-0.4/platform/numpy'
# Change default path when looking for libs to fake dir,
# so we can set everything by env variables
cd work/main.d/numpy-1.0.3.1 && 
PYTHONPATH=/home/cohen/garnumpyinstall/lib/python2.5/site-packages:/home/cohen/garnumpyinstall/lib/python2.5/site-packages/gtk-2.0 
/usr/bin/python \
        setup.py config_fc --fcompiler=gnu config
Traceback (most recent call last):
  File "setup.py", line 90, in <module>
    setup_package()
  File "setup.py", line 60, in setup_package
    from numpy.distutils.core import setup
  File 
"/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/__init__.py", 
line 39, in <module>
    import core
  File 
"/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/__init__.py", 
line 8, in <module>
    import numerictypes as nt
  File 
"/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/numerictypes.py", 
line 83, in <module>
    from numpy.core.multiarray import typeinfo, ndarray, array, empty, dtype
ImportError: No module named multiarray
make[2]: *** [configure-custom] Error 1
make[2]: Leaving directory 
`/data1/sources/python/garnumpy-0.4/platform/numpy'
make[1]: *** [../../platform/numpy/cookies/main.d/install] Error 2
make[1]: Leaving directory 
`/data1/sources/python/garnumpy-0.4/platform/scipy'
make: *** [imgdep-main] Error 2

Any idea? Your package is anyway a great idea. I would actually love to 
see it work, of course, but also allow for the possibility to point to 
already exisiting source directories for numpy and scipy (for instance 
from a previous svn checkout). I did not read the doc of your package so 
it might actually be there....

anyway, thanks in advance for your help.
Johann


From david at ar.media.kyoto-u.ac.jp  Tue Nov 13 01:10:32 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 13 Nov 2007 15:10:32 +0900
Subject: [SciPy-user] Bundling numpy/scipy with applications
In-Reply-To: <47393D17.9090000@slac.stanford.edu>
References: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>	<47392C92.8020208@ar.media.kyoto-u.ac.jp>
	<47393D17.9090000@slac.stanford.edu>
Message-ID: <47393FD8.7060400@ar.media.kyoto-u.ac.jp>

Johann Cohen-Tanugi wrote:
> David Cournapeau wrote:
>> Michael Hearne wrote:
>>   
>>> I'm creating a Python application that is internal to my organization, 
>>> but will be installed on both Linux and Mac OS X machines.  The 
>>> application depends heavily on a number of "non-pure" modules (those 
>>> that have C/C++/FORTRAN components), like numpy, scipy, gdal, etc.
>>>
>>> What is the most "pythonic" way to bundle these types of modules 
>>> inside my application?
>>>
>>> I've investigated dist_utils and setuptools, and I don't see an easy 
>>> way with those tools to include build instructions for packages built 
>>> with autconf tools.
>>>
>>> Is my only recourse to write a custom install script that calls the 
>>> "configure;make;make install" from a shell?
>>>
>>>     
>> You have two problems: building and packaging. For packaging, autoconf 
>> will not help you much outside providing a coherence between packages, I 
>> think; distutils/setuptools have packaging tools, but I don't know how 
>> good they are.
>>
>> On Linux, the packaging system depends on the distribution: since it is 
>> only internal, the problem is severely simplified, since you are likely 
>> to target only one format (rpm, deb, etc...). On Mac OS X, you have also 
>> tools to package .pkg and .mpkg (which are a set of .pkg), available 
>> freely on apple dev website; I have not used them much, but they look 
>> simple and easy to use for simple command line packages.
>>
>> You could hack something in distutils to call configure and so on, but I 
>> don't think it will be pleasant. Building packages using 
>> distutils/setuptools is easy, and do not need much configuration. So I 
>> think it is more natural to use a top level tool calling 
>> distutils/setuptools than to use distutils as the top level tool. As a 
>> pythonic way, you may consider build tools, such as scons or waf, both 
>> written in python.
>>
>> But frankly, I would not bother too much about pythonic way: since some 
>> tools will be shell based anyway (autotools, packaging tools), and you 
>> don't have to care about windows, using make and shell may actually be 
>> easier.
>>
>> For the build part, you may take a look at my garnumpy package: it is 
>> essentially a set of rules for Gnu Makefiles, and it can build numpy + 
>> scipy, with a configurable set of dependencies (ATLAS, NETLIB 
>> BLAS/LAPACK, fftw, etc....). It can build both distutils and 
>> autotools-based packages.
>>
>> http://www.ar.media.kyoto-u.ac.jp/members/david/archives/garnumpy/garnumpy-0.4.tbz2
>>
>> I used it sucessfully on linux and cygwin, so I would say it should work 
>> on Mac OS X without too much trouble. The only thing which will be 
>> likely to be a pain is fat binaries (Universal). I use it to build a 
>> totally self contained numpy/scipy installation, which is a first step 
>> toward packaging. If you think it can be useful to you, don't hesitate 
>> to ask questions; there is also a bzr archive if you want to have access 
>> to the dev history of the tool
>>
>> cheers,
>>
>> David
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>   
> Hi David,
> I am having a hell of a time with lapack/atlas. 
That's because they are far from trivial to build correctly. As always 
with build problems, it is never complicated, but problems are often 
difficult to track down. To make things worse, some linux distribution 
(including fedora and suse) did include bogus packages for blas and 
lapack. That's why I did this garnumpy thing in the first place: quickly 
build blas/lapack correctly, so that I can test things more easily.
> I already posted the 
> issue I currently have with scipy after following the building doc in 
> scipy web site (see 
> http://projects.scipy.org/pipermail/scipy-user/2007-November/014506.html). 
> Besides, octave build seems to also have serious issues with my 
> lapack/atlas build.
Which distribution are you using ?
>
> So I immediately downloaded youor package, but I seem to have a problem 
> with the numpy download :
> make[3]: Leaving directory 
> `/data1/sources/python/garnumpy-0.4/platform/numpy'
> # Change default path when looking for libs to fake dir,
> # so we can set everything by env variables
> cd work/main.d/numpy-1.0.3.1 && 
> PYTHONPATH=/home/cohen/garnumpyinstall/lib/python2.5/site-packages:/home/cohen/garnumpyinstall/lib/python2.5/site-packages/gtk-2.0 
> /usr/bin/python \
>         setup.py config_fc --fcompiler=gnu config
> Traceback (most recent call last):
>   File "setup.py", line 90, in <module>
>     setup_package()
>   File "setup.py", line 60, in setup_package
>     from numpy.distutils.core import setup
>   File 
> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/__init__.py", 
> line 39, in <module>
>     import core
>   File 
> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/__init__.py", 
> line 8, in <module>
>     import numerictypes as nt
>   File 
> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/numerictypes.py", 
> line 83, in <module>
>     from numpy.core.multiarray import typeinfo, ndarray, array, empty, dtype
> ImportError: No module named multiarray
> make[2]: *** [configure-custom] Error 1
> make[2]: Leaving directory 
> `/data1/sources/python/garnumpy-0.4/platform/numpy'
> make[1]: *** [../../platform/numpy/cookies/main.d/install] Error 2
> make[1]: Leaving directory 
> `/data1/sources/python/garnumpy-0.4/platform/scipy'
> make: *** [imgdep-main] Error 2
>
This is really strange, I have never seen this error. This is what I 
would do:
    - remove garnumpyinstall directory
    - go into bootstrap/lapack, and do make install
    - go into platform/numpy, do make install
At this point, you should have an installed numpy in garnumpyinstall: 
test it (import numpy; numpy.test(level = 9999, verbosity = 9999). If 
this works, then go into platform/scipy. The errors should be clearer.
> Any idea? Your package is anyway a great idea. I would actually love to 
> see it work, of course, but also allow for the possibility to point to 
> already exisiting source directories for numpy and scipy (for instance 
> from a previous svn checkout). I did not read the doc of your package so 
> it might actually be there....
You cannot use svn, because this prevents patching from working, and 
also there is a checksum when downloading: this is a major limitation 
(but not so major if you realize that the only thing you really want to 
use from svn are numpy and scipy; you can reuse blas/lapack/fftw/atlas, 
and that's my main use when testing on different platforms).

You can reuse downloaded archives, though:

make garchive

it will download all the tarballs (you cannot choose a subset, 
unfortunately) and put them into a directory which will be reused 
automatically afterwards. This means that even if you do make clean 
anywhere in the source tree, you won't need to download over and over 
the same things.

cheers,

David



From cohen at slac.stanford.edu  Tue Nov 13 02:12:04 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Mon, 12 Nov 2007 23:12:04 -0800
Subject: [SciPy-user] Bundling numpy/scipy with applications
In-Reply-To: <47393FD8.7060400@ar.media.kyoto-u.ac.jp>
References: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>	<47392C92.8020208@ar.media.kyoto-u.ac.jp>	<47393D17.9090000@slac.stanford.edu>
	<47393FD8.7060400@ar.media.kyoto-u.ac.jp>
Message-ID: <47394E44.6010608@slac.stanford.edu>

David Cournapeau wrote:
> Johann Cohen-Tanugi wrote:
>   
>> David Cournapeau wrote:
>>     
>>> Michael Hearne wrote:
>>>   
>>>       
>>>> I'm creating a Python application that is internal to my organization, 
>>>> but will be installed on both Linux and Mac OS X machines.  The 
>>>> application depends heavily on a number of "non-pure" modules (those 
>>>> that have C/C++/FORTRAN components), like numpy, scipy, gdal, etc.
>>>>
>>>> What is the most "pythonic" way to bundle these types of modules 
>>>> inside my application?
>>>>
>>>> I've investigated dist_utils and setuptools, and I don't see an easy 
>>>> way with those tools to include build instructions for packages built 
>>>> with autconf tools.
>>>>
>>>> Is my only recourse to write a custom install script that calls the 
>>>> "configure;make;make install" from a shell?
>>>>
>>>>     
>>>>         
>>> You have two problems: building and packaging. For packaging, autoconf 
>>> will not help you much outside providing a coherence between packages, I 
>>> think; distutils/setuptools have packaging tools, but I don't know how 
>>> good they are.
>>>
>>> On Linux, the packaging system depends on the distribution: since it is 
>>> only internal, the problem is severely simplified, since you are likely 
>>> to target only one format (rpm, deb, etc...). On Mac OS X, you have also 
>>> tools to package .pkg and .mpkg (which are a set of .pkg), available 
>>> freely on apple dev website; I have not used them much, but they look 
>>> simple and easy to use for simple command line packages.
>>>
>>> You could hack something in distutils to call configure and so on, but I 
>>> don't think it will be pleasant. Building packages using 
>>> distutils/setuptools is easy, and do not need much configuration. So I 
>>> think it is more natural to use a top level tool calling 
>>> distutils/setuptools than to use distutils as the top level tool. As a 
>>> pythonic way, you may consider build tools, such as scons or waf, both 
>>> written in python.
>>>
>>> But frankly, I would not bother too much about pythonic way: since some 
>>> tools will be shell based anyway (autotools, packaging tools), and you 
>>> don't have to care about windows, using make and shell may actually be 
>>> easier.
>>>
>>> For the build part, you may take a look at my garnumpy package: it is 
>>> essentially a set of rules for Gnu Makefiles, and it can build numpy + 
>>> scipy, with a configurable set of dependencies (ATLAS, NETLIB 
>>> BLAS/LAPACK, fftw, etc....). It can build both distutils and 
>>> autotools-based packages.
>>>
>>> http://www.ar.media.kyoto-u.ac.jp/members/david/archives/garnumpy/garnumpy-0.4.tbz2
>>>
>>> I used it sucessfully on linux and cygwin, so I would say it should work 
>>> on Mac OS X without too much trouble. The only thing which will be 
>>> likely to be a pain is fat binaries (Universal). I use it to build a 
>>> totally self contained numpy/scipy installation, which is a first step 
>>> toward packaging. If you think it can be useful to you, don't hesitate 
>>> to ask questions; there is also a bzr archive if you want to have access 
>>> to the dev history of the tool
>>>
>>> cheers,
>>>
>>> David
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>   
>>>       
>> Hi David,
>> I am having a hell of a time with lapack/atlas. 
>>     
> That's because they are far from trivial to build correctly. As always 
> with build problems, it is never complicated, but problems are often 
> difficult to track down. To make things worse, some linux distribution 
> (including fedora and suse) did include bogus packages for blas and 
> lapack. That's why I did this garnumpy thing in the first place: quickly 
> build blas/lapack correctly, so that I can test things more easily.
>   
>> I already posted the 
>> issue I currently have with scipy after following the building doc in 
>> scipy web site (see 
>> http://projects.scipy.org/pipermail/scipy-user/2007-November/014506.html). 
>> Besides, octave build seems to also have serious issues with my 
>> lapack/atlas build.
>>     
> Which distribution are you using ?
>   
>> So I immediately downloaded youor package, but I seem to have a problem 
>> with the numpy download :
>> make[3]: Leaving directory 
>> `/data1/sources/python/garnumpy-0.4/platform/numpy'
>> # Change default path when looking for libs to fake dir,
>> # so we can set everything by env variables
>> cd work/main.d/numpy-1.0.3.1 && 
>> PYTHONPATH=/home/cohen/garnumpyinstall/lib/python2.5/site-packages:/home/cohen/garnumpyinstall/lib/python2.5/site-packages/gtk-2.0 
>> /usr/bin/python \
>>         setup.py config_fc --fcompiler=gnu config
>> Traceback (most recent call last):
>>   File "setup.py", line 90, in <module>
>>     setup_package()
>>   File "setup.py", line 60, in setup_package
>>     from numpy.distutils.core import setup
>>   File 
>> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/__init__.py", 
>> line 39, in <module>
>>     import core
>>   File 
>> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/__init__.py", 
>> line 8, in <module>
>>     import numerictypes as nt
>>   File 
>> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/numerictypes.py", 
>> line 83, in <module>
>>     from numpy.core.multiarray import typeinfo, ndarray, array, empty, dtype
>> ImportError: No module named multiarray
>> make[2]: *** [configure-custom] Error 1
>> make[2]: Leaving directory 
>> `/data1/sources/python/garnumpy-0.4/platform/numpy'
>> make[1]: *** [../../platform/numpy/cookies/main.d/install] Error 2
>> make[1]: Leaving directory 
>> `/data1/sources/python/garnumpy-0.4/platform/scipy'
>> make: *** [imgdep-main] Error 2
>>
>>     
> This is really strange, I have never seen this error. This is what I 
> would do:
>     - remove garnumpyinstall directory
>     - go into bootstrap/lapack, and do make install
>     - go into platform/numpy, do make install
> At this point, you should have an installed numpy in garnumpyinstall: 
> test it (import numpy; numpy.test(level = 9999, verbosity = 9999). If 
> this works, then go into platform/scipy. The errors should be clearer.
>   
do make install in platform/numpy fails exactly in the same way :
[cohen at localhost numpy]$ make install
[===== NOW BUILDING:    numpy-1.0.3.1   =====]
        [fetch] complete for numpy.
        [checksum] complete for numpy.
        [extract] complete for numpy.
        [patch] complete for numpy.
# Change default path when looking for libs to fake dir,
# so we can set everything by env variables
cd work/main.d/numpy-1.0.3.1 && 
PYTHONPATH=/home/cohen/garnumpyinstall/lib/python2.5/site-packages:/home/cohen/garnumpyinstall/lib/python2.5/site-packages/gtk-2.0 
/usr/bin/python \
        setup.py config_fc --fcompiler=gnu config
Traceback (most recent call last):
  File "setup.py", line 90, in <module>
    setup_package()
  File "setup.py", line 60, in setup_package
    from numpy.distutils.core import setup
  File 
"/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/__init__.py", 
line 39, in <module>
    import core
  File 
"/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/__init__.py", 
line 8, in <module>
    import numerictypes as nt
  File 
"/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/numerictypes.py", 
line 83, in <module>
    from numpy.core.multiarray import typeinfo, ndarray, array, empty, dtype
ImportError: No module named multiarray

>> Any idea? Your package is anyway a great idea. I would actually love to 
>> see it work, of course, but also allow for the possibility to point to 
>> already exisiting source directories for numpy and scipy (for instance 
>> from a previous svn checkout). I did not read the doc of your package so 
>> it might actually be there....
>>     
> You cannot use svn, because this prevents patching from working, and 
> also there is a checksum when downloading: this is a major limitation 
> (but not so major if you realize that the only thing you really want to 
> use from svn are numpy and scipy; you can reuse blas/lapack/fftw/atlas, 
> and that's my main use when testing on different platforms).
>
> You can reuse downloaded archives, though:
>
> make garchive
>
> it will download all the tarballs (you cannot choose a subset, 
> unfortunately) and put them into a directory which will be reused 
> automatically afterwards. This means that even if you do make clean 
> anywhere in the source tree, you won't need to download over and over 
> the same things.
>
> cheers,
>
> David
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From david at ar.media.kyoto-u.ac.jp  Tue Nov 13 02:35:40 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 13 Nov 2007 16:35:40 +0900
Subject: [SciPy-user] Bundling numpy/scipy with applications
In-Reply-To: <47394E44.6010608@slac.stanford.edu>
References: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>	<47392C92.8020208@ar.media.kyoto-u.ac.jp>	<47393D17.9090000@slac.stanford.edu>	<47393FD8.7060400@ar.media.kyoto-u.ac.jp>
	<47394E44.6010608@slac.stanford.edu>
Message-ID: <473953CC.7010402@ar.media.kyoto-u.ac.jp>

Johann Cohen-Tanugi wrote:
>
> do make install in platform/numpy fails exactly in the same way :
I  forgot to tell you to do make clean before, otherwise, it indeed has 
no way to fail in a different way :)
> [cohen at localhost numpy]$ make install
> [===== NOW BUILDING:    numpy-1.0.3.1   =====]
>         [fetch] complete for numpy.
>         [checksum] complete for numpy.
>         [extract] complete for numpy.
>         [patch] complete for numpy.
> # Change default path when looking for libs to fake dir,
> # so we can set everything by env variables
> cd work/main.d/numpy-1.0.3.1 && 
> PYTHONPATH=/home/cohen/garnumpyinstall/lib/python2.5/site-packages:/home/cohen/garnumpyinstall/lib/python2.5/site-packages/gtk-2.0 
> /usr/bin/python \
>         setup.py config_fc --fcompiler=gnu config
> Traceback (most recent call last):
>   File "setup.py", line 90, in <module>
>     setup_package()
>   File "setup.py", line 60, in setup_package
>     from numpy.distutils.core import setup
>   File 
> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/__init__.py", 
> line 39, in <module>
>     import core
>   File 
> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/__init__.py", 
> line 8, in <module>
>     import numerictypes as nt
>   File 
> "/data1/sources/python/garnumpy-0.4/platform/numpy/work/main.d/numpy-1.0.3.1/numpy/core/numerictypes.py", 
> line 83, in <module>
>     from numpy.core.multiarray import typeinfo, ndarray, array, empty, dtype
> ImportError: No module named multiarray
I really don't understand this error: you are configuring numpy for 
compilation, so it should not try to import numpy.core (which has no way 
to succeed since it is not built yet). IOW, it is a bootstrap error 
(numpy build system tries to import numpy).

If after make clean, it still does not work, you may try to do:

make clean
make patch

And then go into work/ work/main.d/numpy-1.0.3.1/, and just execute 
python setup.py config there, and give me the errors (you can also give 
them to me in private).

David


From haase at msg.ucsf.edu  Tue Nov 13 07:44:59 2007
From: haase at msg.ucsf.edu (Sebastian Haase)
Date: Tue, 13 Nov 2007 13:44:59 +0100
Subject: [SciPy-user] Bundling numpy/scipy with applications
In-Reply-To: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>
References: <0CDABAC6-8EA6-409E-B389-578864A6407E@usgs.gov>
Message-ID: <bc657ead0711130444p2da4e33cta4482b5879df59b7@mail.gmail.com>

Hi,
This might be a heratic idea (in this context):

But how "bad" is it really, to just build all components,
put them in one place, and distribute them as a  zip/tar-archive ?
PYTHONPATH would need to be adjusted as needed - either in a place
like .bashrc or in 3-line script that would then substitute the
"python"-call.

This is of course least feancy, but might be most effective -
especially if it is only for a releatively small group of people.

Cheers,
Sebastian Haase


On Nov 12, 2007 9:26 PM, Michael Hearne <mhearne at usgs.gov> wrote:
> I'm creating a Python application that is internal to my organization, but
> will be installed on both Linux and Mac OS X machines.  The application
> depends heavily on a number of "non-pure" modules (those that have
> C/C++/FORTRAN components), like numpy, scipy, gdal, etc.
>
> What is the most "pythonic" way to bundle these types of modules inside my
> application?
>
> I've investigated dist_utils and setuptools, and I don't see an easy way
> with those tools to include build instructions for packages built with
> autconf tools.
>
> Is my only recourse to write a custom install script that calls the
> "configure;make;make install" from a shell?
>
> --Mike
>
>
>
> ------------------------------------------------------
> Michael Hearne
> mhearne at usgs.gov
> (303) 273-8620
> USGS National Earthquake Information Center
> 1711 Illinois St. Golden CO 80401
> Senior Software Engineer
> Synergetics, Inc.
> ------------------------------------------------------
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


From sandricionut at yahoo.com  Tue Nov 13 08:17:27 2007
From: sandricionut at yahoo.com (sandric ionut)
Date: Tue, 13 Nov 2007 05:17:27 -0800 (PST)
Subject: [SciPy-user] random gaussian array
Message-ID: <201537.55063.qm@web51307.mail.re2.yahoo.com>

Thank you Matthieu:

I looked at numpy.random, but all I have is normalvariate. Can you give me an example?

Ionut

----- Original Message ----
From: Matthieu Brucher <matthieu.brucher at gmail.com>
To: SciPy Users List <scipy-user at scipy.org>
Sent: Monday, November 12, 2007 4:11:08 PM
Subject: Re: [SciPy-user] random gaussian array

Yes, look into numpy.random -> norm

Matthieu


2007/11/12, sandric ionut <sandricionut at yahoo.com>:
Hi:
 
I want to create an random Gaussian 2D array with an exact size. For example I want my array to look like:
 
2 3 4 0 1 1
2 3 3 4 1 1
3 2 5 6 7 8
 
all the values should be randomly generated from a Gaussian distribution with mean and standard deviation
 
Is it possible to do it with scipy or numpy?
 
Thank you
 
Ionut

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org 
http://projects.scipy.org/mailman/listinfo/scipy-user





-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher


      ____________________________________________________________________________________
Get easy, one-click access to your favorites. 
Make Yahoo! your homepage.
http://www.yahoo.com/r/hs 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071113/99380bbb/attachment.html>

From sandricionut at yahoo.com  Tue Nov 13 08:23:29 2007
From: sandricionut at yahoo.com (sandric ionut)
Date: Tue, 13 Nov 2007 05:23:29 -0800 (PST)
Subject: [SciPy-user] random gaussian array SOLVED
Message-ID: <757507.90289.qm@web51301.mail.re2.yahoo.com>

I have found it.

randomArray = numpy.random.normal(mean, stdev, (shape))

It works OK

Thank's again


----- Original Message ----
From: Matthieu Brucher <matthieu.brucher at gmail.com>
To: SciPy Users List <scipy-user at scipy.org>
Sent: Monday, November 12, 2007 4:11:08 PM
Subject: Re: [SciPy-user] random gaussian array

Yes, look into numpy.random -> norm

Matthieu


2007/11/12, sandric ionut <sandricionut at yahoo.com>:
Hi:
 
I want to create an random Gaussian 2D array with an exact size. For example I want my array to look like:
 
2 3 4 0 1 1
2 3 3 4 1 1
3 2 5 6 7 8
 
all the values should be randomly generated from a Gaussian distribution with mean and standard deviation
 
Is it possible to do it with scipy or numpy?
 
Thank you
 
Ionut

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org 
http://projects.scipy.org/mailman/listinfo/scipy-user





-- 
French PhD student
Website : http://miles.developpez.com/
Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher


      ____________________________________________________________________________________
Be a better pen pal. 
Text or chat with friends inside Yahoo! Mail. See how.  http://overview.mail.yahoo.com/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071113/da75846b/attachment.html>

From david at ar.media.kyoto-u.ac.jp  Tue Nov 13 08:13:24 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 13 Nov 2007 22:13:24 +0900
Subject: [SciPy-user] random gaussian array
In-Reply-To: <201537.55063.qm@web51307.mail.re2.yahoo.com>
References: <201537.55063.qm@web51307.mail.re2.yahoo.com>
Message-ID: <4739A2F4.4060709@ar.media.kyoto-u.ac.jp>

sandric ionut wrote:
> Thank you Matthieu:
>  
> I looked at numpy.random, but all I have is normalvariate. Can you 
> give me an example?
This is strange. Does:

 >> from numpy.random import normal

fails ? If you use ipython, you can use:

 >> help numpy.random

To get a list of all the function exporteed by numpy.random. Otherwise, 
with a standard python shell, you can use

 >> print dir(numpy.random)

In both cases, you should see normal as a function.

cheers,

David


From stefan at sun.ac.za  Tue Nov 13 05:04:08 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Tue, 13 Nov 2007 12:04:08 +0200
Subject: [SciPy-user] KDE question
In-Reply-To: <473566E4.2060101@gmail.com>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<4734BF4E.1090102@gmail.com>
	<47356362.7010508@ar.media.kyoto-u.ac.jp>
	<473566E4.2060101@gmail.com>
Message-ID: <20071113100408.GA6802@mentat.za.net>

On Sat, Nov 10, 2007 at 02:08:04AM -0600, Robert Kern wrote:
> David Cournapeau wrote:
> 
> > I am not sure I understand exactly the problem, but if the problem is to 
> > find a contour level of a Gaussian density, it has a closed form for any 
> > dimension.
> 
> No, the problem is to find the appropriate contour level of a kernel density
> estimate (with Gaussian kernels in this case). Essentially, a mixture of many
> Gaussians, not a single one.

Sounds like the kind of problem that can be solved using marching
squares:

http://www.polytech.unice.fr/~lingrand/MarchingCubes/algo.html

Regards
St?fan


From aisaac at american.edu  Tue Nov 13 08:35:26 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Tue, 13 Nov 2007 08:35:26 -0500
Subject: [SciPy-user] random gaussian array
In-Reply-To: <201537.55063.qm@web51307.mail.re2.yahoo.com>
References: <201537.55063.qm@web51307.mail.re2.yahoo.com>
Message-ID: <Mahogany-0.67.0-1696-20071113-083526.00@american.edu>

On Tue, 13 Nov 2007, sandric ionut apparently wrote:
> I looked at numpy.random, but all I have is normalvariate 

But that matches your question?
If you use ``normal``, just set the size (i.e., shape).

Cheers,
Alan Isaac


>>> help(N.random)
Help on package numpy.random in numpy:
...
    multivariate_normal(...)
        Return an array containing multivariate normally distributed random numbers
        with specified mean and covariance.

        multivariate_normal(mean, cov) -> random values
        multivariate_normal(mean, cov, [m, n, ...]) -> random values

        mean must be a 1 dimensional array. cov must be a square two dimensional
        array with the same number of rows and columns as mean has elements.

        The first form returns a single 1-D array containing a multivariate
        normal.

        The second form returns an array of shape (m, n, ..., cov.shape[0]).
        In this case, output[i,j,...,:] is a 1-D array containing a multivariate
        normal.

    normal(...)
        Normal distribution (mean=loc, stdev=scale).

        normal(loc=0.0, scale=1.0, size=None) -> random values




From jeremit0 at gmail.com  Tue Nov 13 09:20:46 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Tue, 13 Nov 2007 09:20:46 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <6ce0ac130711122059x610214bey5698dde00ee9cd8c@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
	<4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
	<3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>
	<6ce0ac130711122059x610214bey5698dde00ee9cd8c@mail.gmail.com>
Message-ID: <3db594f70711130620h5c6914dcq4750ef9111d49a6b@mail.gmail.com>

On Nov 12, 2007 11:59 PM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> Great, now you are ready for the other problem on Leopard:
>
> sudo does not preserve the PYTHONPATH environment variable.
>
> Thus anytime you do::
>
> sudo python setup.py install
>
> to install something else, your version of numpy won't be used!  This
> has been reported to apple and the problem can be fixed by adding
> PYTHONPATH to the appropriate section in /etc/sudoers.
>
> Another option (see the recent thread on pythonmax-sig) is to create a
> .pth file that dynamically reorders the sys.path.  This approach is
> nice as it doesn't require setting PYTHONPATH.
>
> You will run into this when installing matplotlib and pytables...if
> you go that way.
>

I am heading that way, at least I am installing matplotlib.  Now I run
into other issues not yet related to the PYTHONPATH problem.  when
issuing the command


$ python setup.py build

Everything seems to proceed normally until I get the following error:
?
src/_image.cpp:5:17: src/_image.cpp:5:17: error: png.h: No such file
or directory
error: png.h: No such file or directory
src/_image.cpp: In member function 'Py::Object Image::write_png(const
Py::Tuple&)':
src/_image.cpp:646: error: 'png_structp' was not declared in this scope
src/_image.cpp:646: error: expected `;' before 'png_ptr'
?

After that there are many compile errors.  At the beginning of the
building there was a message

libpng: found, but unknown version (no pkg-config)
                        * Could not find 'libpng' headers in any of
                        * '/usr/local/include', '/usr/include', '.'

So it seems like it doesn't know where to find the header files for
libpng.  How can I tell setup where to go looking for the header
files?  (I also got a similar message for freetype2.)

Thanks,
Jeremy


From ellisonbg.net at gmail.com  Tue Nov 13 09:56:12 2007
From: ellisonbg.net at gmail.com (Brian Granger)
Date: Tue, 13 Nov 2007 07:56:12 -0700
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <3db594f70711130620h5c6914dcq4750ef9111d49a6b@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
	<4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
	<3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>
	<6ce0ac130711122059x610214bey5698dde00ee9cd8c@mail.gmail.com>
	<3db594f70711130620h5c6914dcq4750ef9111d49a6b@mail.gmail.com>
Message-ID: <6ce0ac130711130656u5ce11908mec651a17b9c53ed9@mail.gmail.com>

How did you install libpng and freetype? What ---prefix did you use?

Brian

On Nov 13, 2007 7:20 AM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> On Nov 12, 2007 11:59 PM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> > Great, now you are ready for the other problem on Leopard:
> >
> > sudo does not preserve the PYTHONPATH environment variable.
> >
> > Thus anytime you do::
> >
> > sudo python setup.py install
> >
> > to install something else, your version of numpy won't be used!  This
> > has been reported to apple and the problem can be fixed by adding
> > PYTHONPATH to the appropriate section in /etc/sudoers.
> >
> > Another option (see the recent thread on pythonmax-sig) is to create a
> > .pth file that dynamically reorders the sys.path.  This approach is
> > nice as it doesn't require setting PYTHONPATH.
> >
> > You will run into this when installing matplotlib and pytables...if
> > you go that way.
> >
>
> I am heading that way, at least I am installing matplotlib.  Now I run
> into other issues not yet related to the PYTHONPATH problem.  when
> issuing the command
>
>
> $ python setup.py build
>
> Everything seems to proceed normally until I get the following error:
> ?
> src/_image.cpp:5:17: src/_image.cpp:5:17: error: png.h: No such file
> or directory
> error: png.h: No such file or directory
> src/_image.cpp: In member function 'Py::Object Image::write_png(const
> Py::Tuple&)':
> src/_image.cpp:646: error: 'png_structp' was not declared in this scope
> src/_image.cpp:646: error: expected `;' before 'png_ptr'
> ?
>
> After that there are many compile errors.  At the beginning of the
> building there was a message
>
> libpng: found, but unknown version (no pkg-config)
>                         * Could not find 'libpng' headers in any of
>                         * '/usr/local/include', '/usr/include', '.'
>
> So it seems like it doesn't know where to find the header files for
> libpng.  How can I tell setup where to go looking for the header
> files?  (I also got a similar message for freetype2.)
>
> Thanks,
>
> Jeremy
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From jeremit0 at gmail.com  Tue Nov 13 11:33:27 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Tue, 13 Nov 2007 11:33:27 -0500
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <6ce0ac130711130656u5ce11908mec651a17b9c53ed9@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
	<4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
	<3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>
	<6ce0ac130711122059x610214bey5698dde00ee9cd8c@mail.gmail.com>
	<3db594f70711130620h5c6914dcq4750ef9111d49a6b@mail.gmail.com>
	<6ce0ac130711130656u5ce11908mec651a17b9c53ed9@mail.gmail.com>
Message-ID: <3db594f70711130833j7e687137g35099eeec5b701b@mail.gmail.com>

I didn't install them.  I was guessing maybe they came with Leopard.
Using Spotlight I have found header files for libpng and freetype
here:

/Developer/SDKs/MacOSX10.5.sdk/usr/X11/include

Can I use these header files?

Thanks,
Jeremy

On Nov 13, 2007 9:56 AM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> How did you install libpng and freetype? What ---prefix did you use?
>
> Brian
>
>
> On Nov 13, 2007 7:20 AM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> > On Nov 12, 2007 11:59 PM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> > > Great, now you are ready for the other problem on Leopard:
> > >
> > > sudo does not preserve the PYTHONPATH environment variable.
> > >
> > > Thus anytime you do::
> > >
> > > sudo python setup.py install
> > >
> > > to install something else, your version of numpy won't be used!  This
> > > has been reported to apple and the problem can be fixed by adding
> > > PYTHONPATH to the appropriate section in /etc/sudoers.
> > >
> > > Another option (see the recent thread on pythonmax-sig) is to create a
> > > .pth file that dynamically reorders the sys.path.  This approach is
> > > nice as it doesn't require setting PYTHONPATH.
> > >
> > > You will run into this when installing matplotlib and pytables...if
> > > you go that way.
> > >
> >
> > I am heading that way, at least I am installing matplotlib.  Now I run
> > into other issues not yet related to the PYTHONPATH problem.  when
> > issuing the command
> >
> >
> > $ python setup.py build
> >
> > Everything seems to proceed normally until I get the following error:
> > ?
> > src/_image.cpp:5:17: src/_image.cpp:5:17: error: png.h: No such file
> > or directory
> > error: png.h: No such file or directory
> > src/_image.cpp: In member function 'Py::Object Image::write_png(const
> > Py::Tuple&)':
> > src/_image.cpp:646: error: 'png_structp' was not declared in this scope
> > src/_image.cpp:646: error: expected `;' before 'png_ptr'
> > ?
> >
> > After that there are many compile errors.  At the beginning of the
> > building there was a message
> >
> > libpng: found, but unknown version (no pkg-config)
> >                         * Could not find 'libpng' headers in any of
> >                         * '/usr/local/include', '/usr/include', '.'
> >
> > So it seems like it doesn't know where to find the header files for
> > libpng.  How can I tell setup where to go looking for the header
> > files?  (I also got a similar message for freetype2.)
> >
> > Thanks,
> >
> > Jeremy
>
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From ellisonbg.net at gmail.com  Tue Nov 13 11:37:41 2007
From: ellisonbg.net at gmail.com (Brian Granger)
Date: Tue, 13 Nov 2007 09:37:41 -0700
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <3db594f70711130833j7e687137g35099eeec5b701b@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
	<4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
	<3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>
	<6ce0ac130711122059x610214bey5698dde00ee9cd8c@mail.gmail.com>
	<3db594f70711130620h5c6914dcq4750ef9111d49a6b@mail.gmail.com>
	<6ce0ac130711130656u5ce11908mec651a17b9c53ed9@mail.gmail.com>
	<3db594f70711130833j7e687137g35099eeec5b701b@mail.gmail.com>
Message-ID: <6ce0ac130711130837p4fdca00cq6fce50c163289d9d@mail.gmail.com>

I don't think so - at least, I have never had luck using these
versions.  I would ask on the matplotlib list about this issue.  If
they are really there, it would be really nice if the matplotlib build
system could find and use the Apple versions.

In the meantime, if you simply download libpng and freetype2 and then do:

./configure --prefix=/usr/local
make
sudo make install

In the source code directories of both, they will be installed in a
way that matplotlib can find.  From there, the TKAgg and WxAgg
matplotlib backends should run fine.

Brian

On Nov 13, 2007 9:33 AM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> I didn't install them.  I was guessing maybe they came with Leopard.
> Using Spotlight I have found header files for libpng and freetype
> here:
>
> /Developer/SDKs/MacOSX10.5.sdk/usr/X11/include
>
> Can I use these header files?
>
> Thanks,
> Jeremy
>
>
> On Nov 13, 2007 9:56 AM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> > How did you install libpng and freetype? What ---prefix did you use?
> >
> > Brian
> >
> >
> > On Nov 13, 2007 7:20 AM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> > > On Nov 12, 2007 11:59 PM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> > > > Great, now you are ready for the other problem on Leopard:
> > > >
> > > > sudo does not preserve the PYTHONPATH environment variable.
> > > >
> > > > Thus anytime you do::
> > > >
> > > > sudo python setup.py install
> > > >
> > > > to install something else, your version of numpy won't be used!  This
> > > > has been reported to apple and the problem can be fixed by adding
> > > > PYTHONPATH to the appropriate section in /etc/sudoers.
> > > >
> > > > Another option (see the recent thread on pythonmax-sig) is to create a
> > > > .pth file that dynamically reorders the sys.path.  This approach is
> > > > nice as it doesn't require setting PYTHONPATH.
> > > >
> > > > You will run into this when installing matplotlib and pytables...if
> > > > you go that way.
> > > >
> > >
> > > I am heading that way, at least I am installing matplotlib.  Now I run
> > > into other issues not yet related to the PYTHONPATH problem.  when
> > > issuing the command
> > >
> > >
> > > $ python setup.py build
> > >
> > > Everything seems to proceed normally until I get the following error:
> > > ?
> > > src/_image.cpp:5:17: src/_image.cpp:5:17: error: png.h: No such file
> > > or directory
> > > error: png.h: No such file or directory
> > > src/_image.cpp: In member function 'Py::Object Image::write_png(const
> > > Py::Tuple&)':
> > > src/_image.cpp:646: error: 'png_structp' was not declared in this scope
> > > src/_image.cpp:646: error: expected `;' before 'png_ptr'
> > > ?
> > >
> > > After that there are many compile errors.  At the beginning of the
> > > building there was a message
> > >
> > > libpng: found, but unknown version (no pkg-config)
> > >                         * Could not find 'libpng' headers in any of
> > >                         * '/usr/local/include', '/usr/include', '.'
> > >
> > > So it seems like it doesn't know where to find the header files for
> > > libpng.  How can I tell setup where to go looking for the header
> > > files?  (I also got a similar message for freetype2.)
> > >
> > > Thanks,
> > >
> > > Jeremy
> >
> > > _______________________________________________
> > > SciPy-user mailing list
> > > SciPy-user at scipy.org
> > > http://projects.scipy.org/mailman/listinfo/scipy-user
> > >
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From robert.kern at gmail.com  Tue Nov 13 12:57:20 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 13 Nov 2007 11:57:20 -0600
Subject: [SciPy-user] KDE question
In-Reply-To: <20071113100408.GA6802@mentat.za.net>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>	<4734BF4E.1090102@gmail.com>	<47356362.7010508@ar.media.kyoto-u.ac.jp>	<473566E4.2060101@gmail.com>
	<20071113100408.GA6802@mentat.za.net>
Message-ID: <4739E580.9050208@gmail.com>

Stefan van der Walt wrote:
> On Sat, Nov 10, 2007 at 02:08:04AM -0600, Robert Kern wrote:
>> David Cournapeau wrote:
>>
>>> I am not sure I understand exactly the problem, but if the problem is to 
>>> find a contour level of a Gaussian density, it has a closed form for any 
>>> dimension.
>> No, the problem is to find the appropriate contour level of a kernel density
>> estimate (with Gaussian kernels in this case). Essentially, a mixture of many
>> Gaussians, not a single one.
> 
> Sounds like the kind of problem that can be solved using marching
> squares:
> 
> http://www.polytech.unice.fr/~lingrand/MarchingCubes/algo.html

This solves the already-matplotlib-solved problem of drawing the contours given
a level. That still leaves finding the correct level. Or am I underestimating
the potential to reformulate marching squares to solve the integration problem, too?

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From gnchen at gmail.com  Tue Nov 13 13:52:09 2007
From: gnchen at gmail.com (Gen-Nan Chen)
Date: Tue, 13 Nov 2007 10:52:09 -0800
Subject: [SciPy-user] fortran compiler not found Mac Leopard 10.5
In-Reply-To: <6ce0ac130711130837p4fdca00cq6fce50c163289d9d@mail.gmail.com>
References: <3db594f70711120906n11ddd08fkd1fca36b12baa611@mail.gmail.com>
	<6ce0ac130711121039y47445aabu22acaa50ff402f83@mail.gmail.com>
	<3db594f70711121125p1cea3dfaqd11133fd5f24e01a@mail.gmail.com>
	<4D4B58BA-70E1-46E1-A955-7C516BB4BB79@arachnedesign.net>
	<3db594f70711121351w4a803df1u58fadcb2488efdbd@mail.gmail.com>
	<6ce0ac130711122059x610214bey5698dde00ee9cd8c@mail.gmail.com>
	<3db594f70711130620h5c6914dcq4750ef9111d49a6b@mail.gmail.com>
	<6ce0ac130711130656u5ce11908mec651a17b9c53ed9@mail.gmail.com>
	<3db594f70711130833j7e687137g35099eeec5b701b@mail.gmail.com>
	<6ce0ac130711130837p4fdca00cq6fce50c163289d9d@mail.gmail.com>
Message-ID: <588944e60711131052l68db1a40i90a0aceb39d3559b@mail.gmail.com>

I just used i-installer to install everything in /usr/local/.
Gen

On Nov 13, 2007 8:37 AM, Brian Granger <ellisonbg.net at gmail.com> wrote:

> I don't think so - at least, I have never had luck using these
> versions.  I would ask on the matplotlib list about this issue.  If
> they are really there, it would be really nice if the matplotlib build
> system could find and use the Apple versions.
>
> In the meantime, if you simply download libpng and freetype2 and then do:
>
> ./configure --prefix=/usr/local
> make
> sudo make install
>
> In the source code directories of both, they will be installed in a
> way that matplotlib can find.  From there, the TKAgg and WxAgg
> matplotlib backends should run fine.
>
> Brian
>
> On Nov 13, 2007 9:33 AM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> > I didn't install them.  I was guessing maybe they came with Leopard.
> > Using Spotlight I have found header files for libpng and freetype
> > here:
> >
> > /Developer/SDKs/MacOSX10.5.sdk/usr/X11/include
> >
> > Can I use these header files?
> >
> > Thanks,
> > Jeremy
> >
> >
> > On Nov 13, 2007 9:56 AM, Brian Granger <ellisonbg.net at gmail.com> wrote:
> > > How did you install libpng and freetype? What ---prefix did you use?
> > >
> > > Brian
> > >
> > >
> > > On Nov 13, 2007 7:20 AM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> > > > On Nov 12, 2007 11:59 PM, Brian Granger <ellisonbg.net at gmail.com>
> wrote:
> > > > > Great, now you are ready for the other problem on Leopard:
> > > > >
> > > > > sudo does not preserve the PYTHONPATH environment variable.
> > > > >
> > > > > Thus anytime you do::
> > > > >
> > > > > sudo python setup.py install
> > > > >
> > > > > to install something else, your version of numpy won't be used!
>  This
> > > > > has been reported to apple and the problem can be fixed by adding
> > > > > PYTHONPATH to the appropriate section in /etc/sudoers.
> > > > >
> > > > > Another option (see the recent thread on pythonmax-sig) is to
> create a
> > > > > .pth file that dynamically reorders the sys.path.  This approach
> is
> > > > > nice as it doesn't require setting PYTHONPATH.
> > > > >
> > > > > You will run into this when installing matplotlib and
> pytables...if
> > > > > you go that way.
> > > > >
> > > >
> > > > I am heading that way, at least I am installing matplotlib.  Now I
> run
> > > > into other issues not yet related to the PYTHONPATH problem.  when
> > > > issuing the command
> > > >
> > > >
> > > > $ python setup.py build
> > > >
> > > > Everything seems to proceed normally until I get the following
> error:
> > > > ?
> > > > src/_image.cpp:5:17: src/_image.cpp:5:17: error: png.h: No such file
> > > > or directory
> > > > error: png.h: No such file or directory
> > > > src/_image.cpp: In member function 'Py::Object
> Image::write_png(const
> > > > Py::Tuple&)':
> > > > src/_image.cpp:646: error: 'png_structp' was not declared in this
> scope
> > > > src/_image.cpp:646: error: expected `;' before 'png_ptr'
> > > > ?
> > > >
> > > > After that there are many compile errors.  At the beginning of the
> > > > building there was a message
> > > >
> > > > libpng: found, but unknown version (no pkg-config)
> > > >                         * Could not find 'libpng' headers in any of
> > > >                         * '/usr/local/include', '/usr/include', '.'
> > > >
> > > > So it seems like it doesn't know where to find the header files for
> > > > libpng.  How can I tell setup where to go looking for the header
> > > > files?  (I also got a similar message for freetype2.)
> > > >
> > > > Thanks,
> > > >
> > > > Jeremy
> > >
> > > > _______________________________________________
> > > > SciPy-user mailing list
> > > > SciPy-user at scipy.org
> > > > http://projects.scipy.org/mailman/listinfo/scipy-user
> > > >
> > > _______________________________________________
> > > SciPy-user mailing list
> > > SciPy-user at scipy.org
> > > http://projects.scipy.org/mailman/listinfo/scipy-user
> > >
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
Gen-Nan Chen, Ph. D.
email: gnchen at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071113/7f85949d/attachment.html>

From ivilata at carabos.com  Wed Nov 14 12:08:00 2007
From: ivilata at carabos.com (Ivan Vilata i Balaguer)
Date: Wed, 14 Nov 2007 18:08:00 +0100
Subject: [SciPy-user] [ANN] Release of the first PyTables video
Message-ID: <20071114170800.GD30891@tardis.terramar.selidor.net>

=====================================
 Release of the first PyTables video
=====================================

`Carabos <http://www.carabos.com/>`_ is very proud to announce the
first of a series of videos dedicated to introducing the main features
of PyTables to the public in a visual and easy to grasp manner.

  http://www.carabos.com/videos/pytables-1-intro

`PyTables <http://www.pytables.org/>`_ is a Free/Open Source package
designed to handle massive amounts of data in a simple, but highly
efficient way, using the HDF5 file format and NumPy data containers.

This first video is an introductory overview of PyTables, covering the
following topics:

  * HDF5 file creation
  * the object tree
  * homogeneous array storage
  * natural naming
  * working with attributes

With a running length of little more than 10 minutes, you may sit back
and watch it during any short break.

More videos about PyTables will be published in the near future.  Stay
tuned on www.pytables.org for the announcement of the new videos.

We would like to hear your opinion on the video so we can do it better
the next time.  We are also open to suggestions for the topics of
future videos.

Best regards,

::

	Ivan Vilata i Balaguer   >qo<   http://www.carabos.com/
	       C?rabos Coop. V.  V  V   Enjoy Data
	                          ""
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 307 bytes
Desc: Digital signature
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071114/d8d53588/attachment.sig>

From hugues.salamin at gmail.com  Wed Nov 14 12:13:32 2007
From: hugues.salamin at gmail.com (Hugues Salamin)
Date: Wed, 14 Nov 2007 18:13:32 +0100
Subject: [SciPy-user] coercion error when using stats.distirbutions.entropy
Message-ID: <deed985b0711140913y614b9b0fs964bd78bf169e7f5@mail.gmail.com>

Hi,

I come against a strange error when I compute the entropy of discrete
random variable. Here is a small script
showing this problem (python 2.4.4, scipy 0.5.2 and numpy 1.0.1):

[code]
#!/usr/bin/env python2.4

from scipy import stats

val = ( (1,2,3),(0.5,0.25,0.25))
dist = stats.distributions.rv_discrete(name="temp", values = val)
dist.pmf(2)
print dist.entropy()
[/code]

I code the following exception:

Traceback (most recent call last):
  File "./error.py", line 8, in ?
    print dist.entropy()
  File "/usr/lib/python2.4/site-packages/scipy/stats/distributions.py",
line 3787, in entropy
    place(output,cond0,self.vecentropy(*goodargs))
  File "/usr/lib/python2.4/site-packages/numpy/lib/function_base.py",
line 856, in __call__
    _res = array(self.ufunc(*args),copy=False).astype(self.otypes[0])
TypeError: function not supported for these types, and can't coerce
safely to supported types


If a remove the call to pmf, the error disappears. I tried to look
under the hood what was happening but I'm totally at lost,

Any help appreciated

Hugues


From cohen at slac.stanford.edu  Wed Nov 14 15:00:35 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Wed, 14 Nov 2007 12:00:35 -0800
Subject: [SciPy-user] coercion error when using
	stats.distirbutions.entropy
In-Reply-To: <deed985b0711140913y614b9b0fs964bd78bf169e7f5@mail.gmail.com>
References: <deed985b0711140913y614b9b0fs964bd78bf169e7f5@mail.gmail.com>
Message-ID: <473B53E3.2020202@slac.stanford.edu>

Hugues Salamin wrote:
> Hi,
>
> I come against a strange error when I compute the entropy of discrete
> random variable. Here is a small script
> showing this problem (python 2.4.4, scipy 0.5.2 and numpy 1.0.1):
>
> [code]
> #!/usr/bin/env python2.4
>
> from scipy import stats
>
> val = ( (1,2,3),(0.5,0.25,0.25))
> dist = stats.distributions.rv_discrete(name="temp", values = val)
> dist.pmf(2)
> print dist.entropy()
> [/code]
>
> I code the following exception:
>
> Traceback (most recent call last):
>   File "./error.py", line 8, in ?
>     print dist.entropy()
>   File "/usr/lib/python2.4/site-packages/scipy/stats/distributions.py",
> line 3787, in entropy
>     place(output,cond0,self.vecentropy(*goodargs))
>   File "/usr/lib/python2.4/site-packages/numpy/lib/function_base.py",
> line 856, in __call__
>     _res = array(self.ufunc(*args),copy=False).astype(self.otypes[0])
> TypeError: function not supported for these types, and can't coerce
> safely to supported types
>
>
> If a remove the call to pmf, the error disappears. I tried to look
> under the hood what was happening but I'm totally at lost,
>
> Any help appreciated
>
> Hugues
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   

Your code snippet works for me with python 2.5, and numpy/scipy 
1.0.3.1/0.6.0
Johann


From sebastien.maret at gmail.com  Wed Nov 14 14:59:34 2007
From: sebastien.maret at gmail.com (Sebastien Maret)
Date: Wed, 14 Nov 2007 14:59:34 -0500
Subject: [SciPy-user] Scipy-0.6.0 fails to build on MacOS 10.4.10 with
	gfortran (gcc4.2.2)
Message-ID: <m2d4uc7g49.fsf@maretmaclt.astro.lsa.umich.edu>

Hello,

The error log is below:

/sw/bin/gfortran -Wall -L/sw/lib build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/scipy/fftpack/_fftpackmodule.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/drfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zrfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfftnd.o build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/fortranobject.o /sw/lib/djbfft.a -L/sw/lib -L/sw/lib/gcc4.2/lib/gcc/i686-apple-darwin8/4.2.2 -Lbuild/temp.macosx-10.4-i386-2.5 -ldfftpack -lfftw3 -lgfortran -o build/lib.macosx-10.4-i386-2.5/scipy/fftpack/_fftpack.so
/usr/bin/ld: Undefined symbols:
_PyArg_ParseTupleAndKeywords
_PyCObject_AsVoidPtr
_PyCObject_Type
_PyComplex_Type
_PyDict_SetItemString
_PyErr_Clear
_PyErr_Format
_PyErr_NewException
_PyErr_Occurred
_PyErr_Print
_PyErr_SetString
_PyExc_ImportError
_PyExc_RuntimeError
_PyImport_ImportModule
_PyInt_Type
_PyModule_GetDict
_PyNumber_Int
_PyObject_GetAttrString
_PySequence_Check
_PySequence_GetItem
_PyString_FromString
_PyString_Type
_PyType_IsSubtype
_PyType_Type
_Py_BuildValue
_Py_InitModule4
__Py_NoneStruct
_PyCObject_FromVoidPtr
_PyDict_DelItemString
_PyDict_GetItemString
_PyDict_New
_PyExc_AttributeError
_PyExc_TypeError
_PyExc_ValueError
_PyMem_Free
_PyObject_Str
_PyObject_Type
_PyString_AsString
_PyString_ConcatAndDel
_Py_FindMethod
__PyObject_New
_MAIN__
collect2: ld returned 1 exit status
/usr/bin/ld: Undefined symbols:
_PyArg_ParseTupleAndKeywords
_PyCObject_AsVoidPtr
_PyCObject_Type
_PyComplex_Type
_PyDict_SetItemString
_PyErr_Clear
_PyErr_Format
_PyErr_NewException
_PyErr_Occurred
_PyErr_Print
_PyErr_SetString
_PyExc_ImportError
_PyExc_RuntimeError
_PyImport_ImportModule
_PyInt_Type
_PyModule_GetDict
_PyNumber_Int
_PyObject_GetAttrString
_PySequence_Check
_PySequence_GetItem
_PyString_FromString
_PyString_Type
_PyType_IsSubtype
_PyType_Type
_Py_BuildValue
_Py_InitModule4
__Py_NoneStruct
_PyCObject_FromVoidPtr
_PyDict_DelItemString
_PyDict_GetItemString
_PyDict_New
_PyExc_AttributeError
_PyExc_TypeError
_PyExc_ValueError
_PyMem_Free
_PyObject_Str
_PyObject_Type
_PyString_AsString
_PyString_ConcatAndDel
_Py_FindMethod
__PyObject_New
_MAIN__
collect2: ld returned 1 exit status
error: Command "/sw/bin/gfortran -Wall -L/sw/lib build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/scipy/fftpack/_fftpackmodule.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/drfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zrfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfftnd.o build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/fortranobject.o /sw/lib/djbfft.a -L/sw/lib -L/sw/lib/gcc4.2/lib/gcc/i686-apple-darwin8/4.2.2 -Lbuild/temp.macosx-10.4-i386-2.5 -ldfftpack -lfftw3 -lgfortran -o build/lib.macosx-10.4-i386-2.5/scipy/fftpack/_fftpack.so" failed with exit status 1

Thanks in advance for the help.
S?bastien



From robert.kern at gmail.com  Wed Nov 14 16:29:21 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 14 Nov 2007 15:29:21 -0600
Subject: [SciPy-user] Scipy-0.6.0 fails to build on MacOS 10.4.10 with
 gfortran (gcc4.2.2)
In-Reply-To: <m2d4uc7g49.fsf@maretmaclt.astro.lsa.umich.edu>
References: <m2d4uc7g49.fsf@maretmaclt.astro.lsa.umich.edu>
Message-ID: <473B68B1.4060007@gmail.com>

Sebastien Maret wrote:
> Hello,
> 
> The error log is below:
> 
> /sw/bin/gfortran -Wall -L/sw/lib build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/scipy/fftpack/_fftpackmodule.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/drfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zrfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfftnd.o build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/fortranobject.o /sw/lib/djbfft.a -L/sw/lib -L/sw/lib/gcc4.2/lib/gcc/i686-apple-darwin8/4.2.2 -Lbuild/temp.macosx-10.4-i386-2.5 -ldfftpack -lfftw3 -lgfortran -o build/lib.macosx-10.4-i386-2.5/scipy/fftpack/_fftpack.so

Don't use the LDFLAGS environment variable. This overwrites the link options,
including the ones which provide the libpython2.5 library.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From barrywark at gmail.com  Wed Nov 14 20:26:33 2007
From: barrywark at gmail.com (Barry Wark)
Date: Wed, 14 Nov 2007 17:26:33 -0800
Subject: [SciPy-user] addition to scikits
Message-ID: <cd7634ce0711141726v19977007l64c63d73caf635ee@mail.gmail.com>

I'm sorry to spam everyone's inbox with such a simple question: how do
I add a package to the scikits SVN?

I have a possible addition to the scikits collection (a wrapper for
the Approximate Nearest Neighbor library, a kdtree implementation that
provides nearest neighbor and approximate nearest neighbor searching),
which I've previously announced on the list. I've re-worked the
installation and tests for the wrapper to conform to the scikits
format. Hopefully it will be useful to others and the scikits
collection seems an appropriate place to put it.

barry


From robert.kern at gmail.com  Thu Nov 15 02:33:38 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 15 Nov 2007 01:33:38 -0600
Subject: [SciPy-user] addition to scikits
In-Reply-To: <cd7634ce0711141726v19977007l64c63d73caf635ee@mail.gmail.com>
References: <cd7634ce0711141726v19977007l64c63d73caf635ee@mail.gmail.com>
Message-ID: <473BF652.3080103@gmail.com>

Barry Wark wrote:
> I'm sorry to spam everyone's inbox with such a simple question: how do
> I add a package to the scikits SVN?

I have asked our sysadmin to give you the appropriate privileges. He should
email you about passwords and such shortly.

> I have a possible addition to the scikits collection (a wrapper for
> the Approximate Nearest Neighbor library, a kdtree implementation that
> provides nearest neighbor and approximate nearest neighbor searching),
> which I've previously announced on the list. I've re-worked the
> installation and tests for the wrapper to conform to the scikits
> format. Hopefully it will be useful to others and the scikits
> collection seems an appropriate place to put it.

Excellent. I look forward to it. Thanks.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From david at ar.media.kyoto-u.ac.jp  Thu Nov 15 02:36:50 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Thu, 15 Nov 2007 16:36:50 +0900
Subject: [SciPy-user] Scipy-0.6.0 fails to build on MacOS 10.4.10 with
 gfortran (gcc4.2.2)
In-Reply-To: <473B68B1.4060007@gmail.com>
References: <m2d4uc7g49.fsf@maretmaclt.astro.lsa.umich.edu>
	<473B68B1.4060007@gmail.com>
Message-ID: <473BF712.4070704@ar.media.kyoto-u.ac.jp>

Robert Kern wrote:
> Sebastien Maret wrote:
>> Hello,
>>
>> The error log is below:
>>
>> /sw/bin/gfortran -Wall -L/sw/lib build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/scipy/fftpack/_fftpackmodule.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/drfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zrfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfftnd.o build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/fortranobject.o /sw/lib/djbfft.a -L/sw/lib -L/sw/lib/gcc4.2/lib/gcc/i686-apple-darwin8/4.2.2 -Lbuild/temp.macosx-10.4-i386-2.5 -ldfftpack -lfftw3 -lgfortran -o build/lib.macosx-10.4-i386-2.5/scipy/fftpack/_fftpack.so
>
> Don't use the LDFLAGS environment variable. This overwrites the link options,
> including the ones which provide the libpython2.5 library.
I was wondering, since this is causing so much trouble, wouldn't it be 
easier to just emit a warning when LDFLAGS or CFLAGS is used, and 
disable it ?

David



From robert.kern at gmail.com  Thu Nov 15 03:09:27 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 15 Nov 2007 02:09:27 -0600
Subject: [SciPy-user] Scipy-0.6.0 fails to build on MacOS 10.4.10 with
 gfortran (gcc4.2.2)
In-Reply-To: <473BF712.4070704@ar.media.kyoto-u.ac.jp>
References: <m2d4uc7g49.fsf@maretmaclt.astro.lsa.umich.edu>	<473B68B1.4060007@gmail.com>
	<473BF712.4070704@ar.media.kyoto-u.ac.jp>
Message-ID: <473BFEB7.5070607@gmail.com>

David Cournapeau wrote:
> Robert Kern wrote:
>> Sebastien Maret wrote:
>>> Hello,
>>>
>>> The error log is below:
>>>
>>> /sw/bin/gfortran -Wall -L/sw/lib build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/scipy/fftpack/_fftpackmodule.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/drfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zrfft.o build/temp.macosx-10.4-i386-2.5/scipy/fftpack/src/zfftnd.o build/temp.macosx-10.4-i386-2.5/build/src.macosx-10.4-i386-2.5/fortranobject.o /sw/lib/djbfft.a -L/sw/lib -L/sw/lib/gcc4.2/lib/gcc/i686-apple-darwin8/4.2.2 -Lbuild/temp.macosx-10.4-i386-2.5 -ldfftpack -lfftw3 -lgfortran -o build/lib.macosx-10.4-i386-2.5/scipy/fftpack/_fftpack.so
>> Don't use the LDFLAGS environment variable. This overwrites the link options,
>> including the ones which provide the libpython2.5 library.
> I was wondering, since this is causing so much trouble, wouldn't it be 
> easier to just emit a warning when LDFLAGS or CFLAGS is used, and 
> disable it ?

Not alone, no. It's the only way to override those settings. Particularly in the
case of Fortran compilers, which will often be quite different from the C
compiler used to build Python itself, this is an important feature. If you make
another way to fully override those settings, then you may change the LDFLAGS
behavior.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From Laurence.Viry at imag.fr  Thu Nov 15 03:22:11 2007
From: Laurence.Viry at imag.fr (Laurence Viry)
Date: Thu, 15 Nov 2007 09:22:11 +0100
Subject: [SciPy-user] Compiling scipy with Intel ifort and SCSL on an
 Atix350(IA64) with linux
Message-ID: <473C01B3.8020608@imag.fr>

Hi!

I use Python2.4.4 and numpy-1.0.3.1.

*I would like to use SCSL SGI Library (which is compiled with Intel 
Compiler) and intel compilers
 to install numpy and scipy*

the install script didn't recognize the fc_config (--fcompiler=intele) 
options

"
warning: build_ext: f77_compiler=intele is not available.
warning: build_ext: extension 'numpy.linalg.lapack_lite' has Fortran 
libraries but no Fortran linker found, using default linker
..."
customize IntelItaniumFCompiler
option --fcompiler not recognized

"

#Platform information:
python -c 'import os,sys;print os.name,sys.platform'
*posix linux2*

uname -a
*Linux zephir 2.6.5-7.282-sn2 #1 SMP Tue Aug 29 10:40:40 UTC 2006 ia64 
ia64 ia64 GNU/Linux*

# Information about C,C++,Fortran compilers/linkers

ifort -V
I*ntel(R) Fortran Itanium(R) Compiler for Itanium(R)-based applications
Version 9.0    Build 20050430 Package ID: l_fc_p_9.0.021*
Copyright (C) 1985-2005 Intel Corporation.  All rights reserved.
FOR NON-COMMERCIAL USE ONLY

*Intel(R) C Itanium(R) Compiler for Itanium(R)-based applications
Version 9.0    Build 20050722 Package ID: l_cc_c_9.0.024*
Copyright (C) 1985-2005 Intel Corporation.  All rights reserved.
FOR NON-COMMERCIAL USE ONLY

# Python version:
python -c 'import sys;print sys.version'
*2.4.4 (#1, Nov 13 2007, 12:19:56)
[GCC 3.3.3 (SuSE Linux)]*

# Python Numpy version: *1.0.3.1*

# f2py version: *2_3979*

# I try to use SCSL SGI Library which is optimized on our cluster, 
compiled with intel compilers




Configuration file for numpy and scipy: site.cfg
#

[DEFAULT]
library_dirs = /usr/lib:/usr/local/lib/
include_dirs = /usr/include:/usr/local/include

[blas]
library_dirs = /usr/lib
blas_libs= scs

[lapack]
library_dirs = /usr/lib
lapack_libs = scs


[fftw]
library_dirs=/usr/local/stow/fftw-3.1.2/lib
include_dirs = /usr/local/stow/fftw-3.1.2/includepython setup.py config 
--compiler=intele --fcompiler=intele build_clib --compiler=intele 
--fcompiler=intele build_ext --compiler=intele --fcompiler=intele install
libraries = fftw3


# python command to install
*python setup.py config --compiler=intele --fcompiler=intele build_clib 
--compiler=intele --fcompiler=intele build_ext --compiler=intele 
--fcompiler=intele install*

*The script didn't recognize the fc_config (--fcompiler=intele) options*

"
...
warning: build_ext: f77_compiler=intele is not available.
warning: build_ext: extension 'numpy.linalg.lapack_lite' has Fortran 
libraries but no Fortran linker found, using default linker
..."
customize IntelItaniumFCompiler
option --fcompiler not recognized


Couldn't match compiler version for 'Intel(R) Fortran Itanium(R) 
Compiler for Itanium(R)-based applications\nVersion 9.0    Build 
20050430 Package ID: l_fc_p_9.0.021\nCopyright (C) 1985-2005 Intel 
Corporation.  All rights reserved.\nFOR NON-COMMERCIAL USE ONLY\n\n 
Intel Fortran 9.0-5238'

I changed numpy/distutils/ccompiler.py to :
compiler_class['intele'] = ('intelccompiler','IntelItaniumCCompiler',
                          "Intel(R) C Itanium(R) Compiler for 
Itanium(R)-based applications")

and put version_match = intel_version_match('Itanium') in 
fcompiler/intel.py



It found the C Intel Compiler but don't find the Fortran Intel compiler.

It's Ok for numpy but when I installed scipy, I had the error with fortran

" building 'scipy.interpolate.dfitpack' extension
error: extension 'scipy.interpolate.dfitpack' has Fortran sources but no 
Fortran compiler found
"
Can you help me to understand

thanks

-- 
Laurence Viry

CRIP UJF - Projet MIRAGE ( http://mirage.imag.fr )
Laboratoire de Mod?lisation et Calcul - IMAG

tel: 04 76 63 56 34
fax: 04 76 63 12 63
e-mail: Laurence.Viry at imag.fr



From darkwind.87 at gmail.com  Thu Nov 15 03:56:06 2007
From: darkwind.87 at gmail.com (Dark Wind)
Date: Thu, 15 Nov 2007 00:56:06 -0800
Subject: [SciPy-user] Problem with NLP in OpenOpt
Message-ID: <8a6035a00711150056r62c30b6cyd5fb4052eaa3f8b8@mail.gmail.com>

Hi,

I am using NLP of OpenOpt to solve a NonLinear programming problem. I
am using scipy_cobyla solver.
The initial point that I have provided is feasible, but I am getting a message:
------------------------------------------message---------------------------------------------------------
starting solver scipy_cobyla (license: BSD )  with problem  unnamed
solver scipy_cobyla has finished solving the problem unnamed
istop:  -100
Solver:   Time Elapsed = 0.75 	CPU Time Elapsed = 0.750110139697
NO FEASIBLE SOLUTION is obtained (max residual = 1.29567664921e-005)
-----------------------------------------------------------------------------------------------------------------


All other solvers cannot take h constraint and some others require
extra additions.

Following is my code:
--------------------------------------------------code-------------------------------------------------
from scikits.openopt import NLP

f = lambda x: 1*x[0] + 7*x[5] + 3*x[10] + 8*x[15]
x0 = [1,0,0,0,0,1,0,0,0,0,1,0,0,0,0,1]

p=NLP(f,x0,maxIter=1e7,maxFunEvals=1e7)

h1 = lambda x: x[0]**2 + x[1]**2 + x[2]**2 + x[3]**2 -1
h2 = lambda x: x[0]*x[4]+x[1]*x[5]+x[2]*x[6]+x[3]*x[7]
h3 = lambda x: x[0]*x[8]+x[1]*x[9]+x[2]*x[10]+x[3]*x[11]
h4 = lambda x: x[0]*x[12]+x[1]*x[13]+x[2]*x[14]+x[3]*x[15]
h5 = lambda x: x[4]**2+x[5]**2+x[6]**2+x[7]**2-1
h6 = lambda x: x[4]*x[8]+x[5]*x[9]+x[6]*x[10]+x[7]*x[11]
h7 = lambda x: x[4]*x[12]+x[5]*x[13]+x[6]*x[14]+x[7]*x[15]
h8 = lambda x: x[8]**2+x[9]**2+x[10]**2+x[11]**2-1
h9 = lambda x: x[8]*x[12]+x[9]*x[13]+x[10]*x[14]+x[11]*x[15]
h10 = lambda x: x[12]**2+x[13]**2+x[14]**2+x[15]**2-1

p.h=[h1,h2,h3,h4,h5,h6,h7,h8,h9,h10]

r=p.solve('scipy_cobyla')
-----------------------------------------------------------------------------------------------------------

The initial value x0 satisfies the constraints hi's but i am still
getting that there is no feasible solution.

Can anyone help me with this?

Thank you


From stefan at sun.ac.za  Thu Nov 15 04:46:53 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Thu, 15 Nov 2007 11:46:53 +0200
Subject: [SciPy-user] KDE question
In-Reply-To: <4739E580.9050208@gmail.com>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<4734BF4E.1090102@gmail.com>
	<47356362.7010508@ar.media.kyoto-u.ac.jp>
	<473566E4.2060101@gmail.com> <20071113100408.GA6802@mentat.za.net>
	<4739E580.9050208@gmail.com>
Message-ID: <20071115094653.GD9382@mentat.za.net>

On Tue, Nov 13, 2007 at 11:57:20AM -0600, Robert Kern wrote:
> Stefan van der Walt wrote:
> > On Sat, Nov 10, 2007 at 02:08:04AM -0600, Robert Kern wrote:
> >> David Cournapeau wrote:
> >>
> >>> I am not sure I understand exactly the problem, but if the problem is to 
> >>> find a contour level of a Gaussian density, it has a closed form for any 
> >>> dimension.
> >> No, the problem is to find the appropriate contour level of a kernel density
> >> estimate (with Gaussian kernels in this case). Essentially, a mixture of many
> >> Gaussians, not a single one.
> > 
> > Sounds like the kind of problem that can be solved using marching
> > squares:
> > 
> > http://www.polytech.unice.fr/~lingrand/MarchingCubes/algo.html
> 
> This solves the already-matplotlib-solved problem of drawing the contours given
> a level. That still leaves finding the correct level. Or am I underestimating
> the potential to reformulate marching squares to solve the
> integration problem, too?

No, I don't think you are.  As for the line-search, since the
different components of the mixture are available, can't we evaluate
the integral (over each component) directly, rather than working with
a grid?

I come from a GMM background, so a question about KDE: is there a
component placed around each datapoint?  It looks that way, since
there are no means calculated anywhere, as with gaussian mixtures.

Cheers
St?fan


From stefan at sun.ac.za  Thu Nov 15 08:05:15 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Thu, 15 Nov 2007 15:05:15 +0200
Subject: [SciPy-user] KDE question
In-Reply-To: <20071115094653.GD9382@mentat.za.net>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>
	<4734BF4E.1090102@gmail.com>
	<47356362.7010508@ar.media.kyoto-u.ac.jp>
	<473566E4.2060101@gmail.com> <20071113100408.GA6802@mentat.za.net>
	<4739E580.9050208@gmail.com> <20071115094653.GD9382@mentat.za.net>
Message-ID: <20071115130515.GC26022@mentat.za.net>

On Thu, Nov 15, 2007 at 11:46:53AM +0200, Stefan van der Walt wrote:
> > > Sounds like the kind of problem that can be solved using marching
> > > squares:
> > > 
> > > http://www.polytech.unice.fr/~lingrand/MarchingCubes/algo.html
> > 
> > This solves the already-matplotlib-solved problem of drawing the contours given
> > a level. That still leaves finding the correct level. Or am I underestimating
> > the potential to reformulate marching squares to solve the
> > integration problem, too?
> 
> No, I don't think you are.  As for the line-search, since the
> different components of the mixture are available, can't we evaluate
> the integral (over each component) directly, rather than working with
> a grid?

No, since the relevant area depends on the *sum* of components, not
the value of the component itself.

Cheers
St?fan


From dmitrey.kroshko at scipy.org  Thu Nov 15 09:49:19 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Thu, 15 Nov 2007 16:49:19 +0200
Subject: [SciPy-user] Problem with NLP in OpenOpt
In-Reply-To: <8a6035a00711150056r62c30b6cyd5fb4052eaa3f8b8@mail.gmail.com>
References: <8a6035a00711150056r62c30b6cyd5fb4052eaa3f8b8@mail.gmail.com>
Message-ID: <473C5C6F.9080208@scipy.org>

hi,
scipy_cobyla yields solution r.ff = -19,

r.xf= array([ -1.00000000e+00,  -9.80461510e-08,   1.35489449e-05,
         5.37394643e-06,   9.80515487e-08,  -1.00000000e+00,
         4.43639580e-07,  -2.00656699e-07,  -5.92177803e-07,
        -2.01266552e-07,  -1.00000000e+00,   1.02019241e-07,
        -1.81972968e-07,   2.91935590e-07,  -1.02019165e-07,
        -1.00000000e+00])

and max constraint = 1.29567664921e-05, that exceeds default p.contol=1e-6

ALGENCAN yields r.ff=-19.0000051782,
r.xf = array([ -1.00000000e+00,   3.77969060e-08,   2.50917684e-08,
         3.90370801e-08,  -3.78516336e-08,  -1.00000020e+00,
        -1.99934148e-08,   3.28135157e-09,  -2.50878312e-08,
         1.99952165e-08,  -9.99999998e-01,   2.31707460e-08,
        -3.90372694e-08,  -3.28049018e-09,  -2.27805959e-08,
        -1.00000048e+00])
and max constraint 9.53914e-07 that is less than default 1e-6 and hence 
is feasible, no error message (btw time elapsed is 30% less than cobyla).
You could just set p.contol=1e-4 and scipy_cobyla will solve your 
problem well.
Regards, D.

Dark Wind wrote:
> Hi,
>
> I am using NLP of OpenOpt to solve a NonLinear programming problem. I
> am using scipy_cobyla solver.
> The initial point that I have provided is feasible, but I am getting a message:
> ------------------------------------------message---------------------------------------------------------
> starting solver scipy_cobyla (license: BSD )  with problem  unnamed
> solver scipy_cobyla has finished solving the problem unnamed
> istop:  -100
> Solver:   Time Elapsed = 0.75 	CPU Time Elapsed = 0.750110139697
> NO FEASIBLE SOLUTION is obtained (max residual = 1.29567664921e-005)
> -----------------------------------------------------------------------------------------------------------------
>
>
> All other solvers cannot take h constraint and some others require
> extra additions.
>
> Following is my code:
> --------------------------------------------------code-------------------------------------------------
> from scikits.openopt import NLP
>
> f = lambda x: 1*x[0] + 7*x[5] + 3*x[10] + 8*x[15]
> x0 = [1,0,0,0,0,1,0,0,0,0,1,0,0,0,0,1]
>
> p=NLP(f,x0,maxIter=1e7,maxFunEvals=1e7)
>
> h1 = lambda x: x[0]**2 + x[1]**2 + x[2]**2 + x[3]**2 -1
> h2 = lambda x: x[0]*x[4]+x[1]*x[5]+x[2]*x[6]+x[3]*x[7]
> h3 = lambda x: x[0]*x[8]+x[1]*x[9]+x[2]*x[10]+x[3]*x[11]
> h4 = lambda x: x[0]*x[12]+x[1]*x[13]+x[2]*x[14]+x[3]*x[15]
> h5 = lambda x: x[4]**2+x[5]**2+x[6]**2+x[7]**2-1
> h6 = lambda x: x[4]*x[8]+x[5]*x[9]+x[6]*x[10]+x[7]*x[11]
> h7 = lambda x: x[4]*x[12]+x[5]*x[13]+x[6]*x[14]+x[7]*x[15]
> h8 = lambda x: x[8]**2+x[9]**2+x[10]**2+x[11]**2-1
> h9 = lambda x: x[8]*x[12]+x[9]*x[13]+x[10]*x[14]+x[11]*x[15]
> h10 = lambda x: x[12]**2+x[13]**2+x[14]**2+x[15]**2-1
>
> p.h=[h1,h2,h3,h4,h5,h6,h7,h8,h9,h10]
>
> r=p.solve('scipy_cobyla')
> -----------------------------------------------------------------------------------------------------------
>
> The initial value x0 satisfies the constraints hi's but i am still
> getting that there is no feasible solution.
>
> Can anyone help me with this?
>
> Thank you
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From dominique.orban at gmail.com  Thu Nov 15 11:54:44 2007
From: dominique.orban at gmail.com (Dominique Orban)
Date: Thu, 15 Nov 2007 11:54:44 -0500
Subject: [SciPy-user] Problem with NLP in OpenOpt
In-Reply-To: <8a6035a00711150056r62c30b6cyd5fb4052eaa3f8b8@mail.gmail.com>
References: <8a6035a00711150056r62c30b6cyd5fb4052eaa3f8b8@mail.gmail.com>
Message-ID: <8793ae6e0711150854g7b78db7ex4738dd0dd00cbe13@mail.gmail.com>

On 11/15/07, Dark Wind <darkwind.87 at gmail.com> wrote:
> The initial value x0 satisfies the constraints hi's but i am still
> getting that there is no feasible solution.
>
> Can anyone help me with this?

Cobyla is an infeasible method, meaning that in order to identifiy a
solution to your problem, it will venture outside of the feasible set.
In your example, it wasn't able to reach a feasible solution within
the given limits (max number of iterations, tolerances, ...) It is not
because their is a feasible point that the code will be able to find a
solution; in this case, it got stuck before reaching the feasible set
again.

You can either re-solve your problem with looser tolerances, which is
not very satisfactory, you can re-solve from a different starting
point (which doesn't have to be feasible), or you can try another
solver (as Dmitrey showed in his response).

Dominique


From adam.ginsburg at colorado.edu  Thu Nov 15 14:07:51 2007
From: adam.ginsburg at colorado.edu (Adam Ginsburg)
Date: Thu, 15 Nov 2007 12:07:51 -0700
Subject: [SciPy-user] Pylab plotting: math range errors unrecoverable
Message-ID: <fceb3d4a0711151107t7fe15b1fv4138dc70b61d9b75@mail.gmail.com>

I've been encountering these errors quite often:
<type 'exceptions.OverflowError'>: math range error

> /usr/lib/python2.5/site-packages/matplotlib/ticker.py(791)scale_range()

    790     else:

--> 791         ex = divmod(math.log10(-meanv), 1)[0]

    792         offset = -10**ex


They happen when I try to plot an array containing "inf" values - no
surprise there.  Two questions, though:
1. is there any way to make pylab ignore inf values, i.e. plot zeroes
or quit cleanly instead of giving a messy error?  I've looked at
masked arrays, but they don't seem to do it, and I'd like something
that does not require modifying my data
2. when I get the above error, then try to plot something WITHOUT
infinities, it fails with the exact same error, and when I ran the
debugger it claimed that meanv was still inf.  Is there any way to fix
that short of quitting and reentering the python command line?

Thanks,
Adam


From mforbes at physics.ubc.ca  Thu Nov 15 14:19:30 2007
From: mforbes at physics.ubc.ca (Michael McNeil Forbes)
Date: Thu, 15 Nov 2007 11:19:30 -0800
Subject: [SciPy-user] Pylab plotting: math range errors unrecoverable
In-Reply-To: <fceb3d4a0711151107t7fe15b1fv4138dc70b61d9b75@mail.gmail.com>
References: <fceb3d4a0711151107t7fe15b1fv4138dc70b61d9b75@mail.gmail.com>
Message-ID: <B3E9751A-B518-4ABB-9DB7-D0BFAB7E2290@physics.ubc.ca>

On 15 Nov 2007, at 11:07 AM, Adam Ginsburg wrote:
> 2. when I get the above error, then try to plot something WITHOUT
> infinities, it fails with the exact same error, and when I ran the
> debugger it claimed that meanv was still inf.  Is there any way to fix
> that short of quitting and reentering the python command line?
You can usually just clear the plot with clf().  I think the plot  
keeps the old data and then tries to redisplay it each time, giving  
errors over and over until you clear the figure.

Michael.





From david at ar.media.kyoto-u.ac.jp  Thu Nov 15 22:44:17 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 16 Nov 2007 12:44:17 +0900
Subject: [SciPy-user] KDE question
In-Reply-To: <20071115130515.GC26022@mentat.za.net>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>	<4734BF4E.1090102@gmail.com>	<47356362.7010508@ar.media.kyoto-u.ac.jp>	<473566E4.2060101@gmail.com>
	<20071113100408.GA6802@mentat.za.net>	<4739E580.9050208@gmail.com>
	<20071115094653.GD9382@mentat.za.net>
	<20071115130515.GC26022@mentat.za.net>
Message-ID: <473D1211.70704@ar.media.kyoto-u.ac.jp>

Stefan van der Walt wrote:
> On Thu, Nov 15, 2007 at 11:46:53AM +0200, Stefan van der Walt wrote:
>>>> Sounds like the kind of problem that can be solved using marching
>>>> squares:
>>>>
>>>> http://www.polytech.unice.fr/~lingrand/MarchingCubes/algo.html
>>> This solves the already-matplotlib-solved problem of drawing the contours given
>>> a level. That still leaves finding the correct level. Or am I underestimating
>>> the potential to reformulate marching squares to solve the
>>> integration problem, too?
>> No, I don't think you are.  As for the line-search, since the
>> different components of the mixture are available, can't we evaluate
>> the integral (over each component) directly, rather than working with
>> a grid?
>
> No, since the relevant area depends on the *sum* of components, not
> the value of the component itself.
Yes, that's what I would have thought, too. As Robert said, I don't 
think you can find an analytical version for the inverse cumulative 
"function" of a mixture of Gaussian on an area of interest, and that's 
what we need (matlab does not implement it, and I guess that if it was 
possible, they would have put it with the pdf and cdf abilities of their 
mixture object). For a component, the contour shape is easy to see 
(ellipsoids), for mixtures, not so easy.

For kernel estimators, you assume each component is the same 'shape', (I 
mean same covariance matrix) right ? Maybe this make the computation 
feasible (find all the points such as sum_i{a_i f(x - \mu_i)} = cst) ?

David


From yennifersantiago at gmail.com  Thu Nov 15 22:59:32 2007
From: yennifersantiago at gmail.com (Yennifer Santiago)
Date: Thu, 15 Nov 2007 23:59:32 -0400
Subject: [SciPy-user] Scipy-AlgoritmosGeneticos
Message-ID: <41bc705b0711151959u49c6a6d2r82a8e70414849e8@mail.gmail.com>

Hola,

Por favor necesito informaci?n sobre el uso de Scipy, especificamente
para aplicaciones con Algoritmos Gen?ticos. Si tienen o conocen alguna
direcci?n donde pueda encontrar algun tutorial sobre esto.

Gracias

Yennifer Santiago


From robert.kern at gmail.com  Thu Nov 15 23:03:59 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 15 Nov 2007 22:03:59 -0600
Subject: [SciPy-user] KDE question
In-Reply-To: <473D1211.70704@ar.media.kyoto-u.ac.jp>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>	<4734BF4E.1090102@gmail.com>	<47356362.7010508@ar.media.kyoto-u.ac.jp>	<473566E4.2060101@gmail.com>	<20071113100408.GA6802@mentat.za.net>	<4739E580.9050208@gmail.com>	<20071115094653.GD9382@mentat.za.net>	<20071115130515.GC26022@mentat.za.net>
	<473D1211.70704@ar.media.kyoto-u.ac.jp>
Message-ID: <473D16AF.9010108@gmail.com>

David Cournapeau wrote:

> For kernel estimators, you assume each component is the same 'shape', (I 
> mean same covariance matrix) right ?

This is the only form that has been implemented, yes. Fancier implementations
can adapt the covariance matrix (in KDE jargon, "bandwidth") of the kernels to
take into account the properties of the local neighborhood. However, this is
usually only done in 1-D because N-D sorting is hard.

> Maybe this make the computation 
> feasible (find all the points such as sum_i{a_i f(x - \mu_i)} = cst) ?

That boils down to evaluating on the grid again, I think.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From cohen at slac.stanford.edu  Thu Nov 15 23:04:45 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Thu, 15 Nov 2007 20:04:45 -0800
Subject: [SciPy-user] Scipy-AlgoritmosGeneticos
In-Reply-To: <41bc705b0711151959u49c6a6d2r82a8e70414849e8@mail.gmail.com>
References: <41bc705b0711151959u49c6a6d2r82a8e70414849e8@mail.gmail.com>
Message-ID: <473D16DD.4050303@slac.stanford.edu>

Yennifer Santiago wrote:
> Hola,
>
> Por favor necesito informaci?n sobre el uso de Scipy, especificamente
> para aplicaciones con Algoritmos Gen?ticos. Si tienen o conocen alguna
> direcci?n donde pueda encontrar algun tutorial sobre esto.
>
> Gracias
>
> Yennifer Santiago
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   
I think that in the sandbox area of scipy there is a genetic algorithm 
package in development. No idea about the stage of development.....
Sorry to answer in English!
Johann


From david at ar.media.kyoto-u.ac.jp  Thu Nov 15 23:13:58 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 16 Nov 2007 13:13:58 +0900
Subject: [SciPy-user] KDE question
In-Reply-To: <473D16AF.9010108@gmail.com>
References: <9F9D8106-7EEE-4623-98C9-B8B65C428419@stanford.edu>	<4734BF4E.1090102@gmail.com>	<47356362.7010508@ar.media.kyoto-u.ac.jp>	<473566E4.2060101@gmail.com>	<20071113100408.GA6802@mentat.za.net>	<4739E580.9050208@gmail.com>	<20071115094653.GD9382@mentat.za.net>	<20071115130515.GC26022@mentat.za.net>	<473D1211.70704@ar.media.kyoto-u.ac.jp>
	<473D16AF.9010108@gmail.com>
Message-ID: <473D1906.5090907@ar.media.kyoto-u.ac.jp>

Robert Kern wrote:
> David Cournapeau wrote:
>
>   
>> For kernel estimators, you assume each component is the same 'shape', (I 
>> mean same covariance matrix) right ?
>>     
>
> This is the only form that has been implemented, yes. Fancier implementations
> can adapt the covariance matrix (in KDE jargon, "bandwidth") of the kernels to
> take into account the properties of the local neighborhood. However, this is
> usually only done in 1-D because N-D sorting is hard.
>
>   
>> Maybe this make the computation 
>> feasible (find all the points such as sum_i{a_i f(x - \mu_i)} = cst) ?
>>     
>
> That boils down to evaluating on the grid again, I think.
>   
Yes, sorry, I was stupid. That's the different locations which matters, 
having the same shape for each Gaussian does not help that much. For one 
component case, the thing which makes it easy it that pdf = cst boils 
down to a "quadratic form involving the samples = constant". For 
multiple components, this does not seem feasible anymore.

cheers,

David



From skraelings001 at gmail.com  Thu Nov 15 23:46:08 2007
From: skraelings001 at gmail.com (Reynaldo Baquerizo)
Date: Thu, 15 Nov 2007 23:46:08 -0500
Subject: [SciPy-user] Scipy-AlgoritmosGeneticos
In-Reply-To: <473D16DD.4050303@slac.stanford.edu>
References: <41bc705b0711151959u49c6a6d2r82a8e70414849e8@mail.gmail.com>
	<473D16DD.4050303@slac.stanford.edu>
Message-ID: <473D2090.6090306@gmail.com>

Johann Cohen-Tanugi escribi?:
> Yennifer Santiago wrote:
>   
>> Hola,
>>
>> Por favor necesito informaci?n sobre el uso de Scipy, especificamente
>> para aplicaciones con Algoritmos Gen?ticos. Si tienen o conocen alguna
>> direcci?n donde pueda encontrar algun tutorial sobre esto.
>>
>> Gracias
>>
>> Yennifer Santiago
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>   
>>     
> I think that in the sandbox area of scipy there is a genetic algorithm
>   
maybe you could look at example.py in:

http://svn.scipy.org/svn/scipy/trunk/scipy/sandbox/ga/


some resources:
http://www.iitk.ac.in/kangal/links.shtml
http://www.genetic-programming.org/
> package in development. No idea about the stage of development.....
> Sorry to answer in English!
> Johann
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   



From jelleferinga at gmail.com  Fri Nov 16 04:22:32 2007
From: jelleferinga at gmail.com (jelle)
Date: Fri, 16 Nov 2007 09:22:32 +0000 (UTC)
Subject: [SciPy-user] Scipy-AlgoritmosGeneticos
References: <41bc705b0711151959u49c6a6d2r82a8e70414849e8@mail.gmail.com>
	<473D16DD.4050303@slac.stanford.edu> <473D2090.6090306@gmail.com>
Message-ID: <loom.20071116T092043-800@post.gmane.org>

if your interested in evolutionary computing, have a look at Evolving Objects.
also, python wrappers exist. 
think of EO as a framework for evolutionary computing



From grh at mur.at  Fri Nov 16 05:44:43 2007
From: grh at mur.at (Georg Holzmann)
Date: Fri, 16 Nov 2007 11:44:43 +0100
Subject: [SciPy-user] adaptive simulated annealing (ASA)
Message-ID: <473D749B.6070308@mur.at>

Hallo!

Is there already a python implementation of the adaptive simulated 
annealing optimization algorithm (maybe a wrapper to the following C 
library: http://www.ingber.com/ASA-README.html) ?

I did quite some google search and did not find something - so any hint 
is very much appreciated !

Thanks,
LG
Georg


From dmitrey.kroshko at scipy.org  Fri Nov 16 05:56:43 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Fri, 16 Nov 2007 12:56:43 +0200
Subject: [SciPy-user] adaptive simulated annealing (ASA)
In-Reply-To: <473D749B.6070308@mur.at>
References: <473D749B.6070308@mur.at>
Message-ID: <473D776B.30009@scipy.org>

Georg Holzmann wrote:
> Hallo!
>
> Is there already a python implementation of the adaptive simulated 
> annealing optimization algorithm (maybe a wrapper to the following C 
> library: http://www.ingber.com/ASA-README.html) ?
>
> I did quite some google search and did not find something - so any hint 
> is very much appreciated !
>   
I had did the search as well (about searching global minimum software 
with python interface enable). All I had found were some obsolete 
packages, + one more with very complicated documentation (IIRC particle 
swarm simulation based).

As for scipy anneal, I had informed it doesn't work properly but noone 
answered

http://article.gmane.org/gmane.comp.python.scientific.user/13553/match=anneal

and I have no time for now to search and fix the bug(s) by myself


Regards, D


From lbolla at gmail.com  Fri Nov 16 09:53:00 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Fri, 16 Nov 2007 15:53:00 +0100
Subject: [SciPy-user] ode on complex systems
Message-ID: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>

Hi all,
can I use any of scipy.integrate.ode/odeint etc. for complex systems? How?

Here is what I'm doing, but I only get r.y real, instead of complex (and the
result is different from Matlab).

----------------------------------------

import
numpy

import scipy.integrate.ode



def f(t, y, betaA, betaB,
CAB):

    M = numpy.array([[1j * betaA, CAB], [-CAB, 1j *
betaB]])

    return numpy.dot(M,
y)



betaA =
1.

betaB =
1.

CAB =
1.



t0 =
0.

t1 =
1.

dt =
.01

y0 = numpy.array([1,
0]).astype(complex)



r = scipy.integrate.ode(f).set_integrator('vode').set_initial_value(y0,
t0).set_f_params(betaA, betaB,
CAB)

while r.successful() and r.t <
t1:

    r.integrate(r.t +
dt)

    print r.t, r.y



----------------------------------------

Thank you in advance,
Lorenzo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071116/f19de794/attachment.html>

From Fabrice.Silva at crans.org  Fri Nov 16 10:19:08 2007
From: Fabrice.Silva at crans.org (Fabrice Silva)
Date: Fri, 16 Nov 2007 16:19:08 +0100
Subject: [SciPy-user] ode on complex systems
In-Reply-To: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
References: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
Message-ID: <1195226348.24691.2.camel@localhost>


Le vendredi 16 novembre 2007 ? 15:53 +0100, lorenzo bolla a ?crit :
> Hi all,
> can I use any of scipy.integrate.ode/odeint etc. for complex systems? How?

You may need to separate real and imaginary parts of your complex
unknowns :
you need to transform
y(t)=[y0(t), y1(t)] with y0 and y1 complex signals
into:
y(t)=[real(y0(t)), imag(y0(t)), real (y1(t)), imag(y1(t))]

and modify your matrix M as well.

Can anyone confirm?
-- 
Fabricio




From nwagner at iam.uni-stuttgart.de  Fri Nov 16 10:25:53 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 16 Nov 2007 16:25:53 +0100
Subject: [SciPy-user] ode on complex systems
In-Reply-To: <1195226348.24691.2.camel@localhost>
References: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
	<1195226348.24691.2.camel@localhost>
Message-ID: <web-92145595@uni-stuttgart.de>

On Fri, 16 Nov 2007 16:19:08 +0100
  Fabrice Silva <Fabrice.Silva at crans.org> wrote:
> 
> Le vendredi 16 novembre 2007 ? 15:53 +0100, lorenzo 
>bolla a ?crit :
>> Hi all,
>> can I use any of scipy.integrate.ode/odeint etc. for 
>>complex systems? How?
> 
> You may need to separate real and imaginary parts of 
>your complex
> unknowns :
> you need to transform
> y(t)=[y0(t), y1(t)] with y0 and y1 complex signals
> into:
> y(t)=[real(y0(t)), imag(y0(t)), real (y1(t)), 
>imag(y1(t))]
> 
> and modify your matrix M as well.
> 
> Can anyone confirm?
> -- 
>Fabricio
> 
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

AFAIK, you can also apply the patch which is available
at

http://projects.scipy.org/scipy/scipy/ticket/334

  
So you can integrate complex ODE's directly without
doubling the size. Please let me know how it works
for you.

Nils



From bryanv at enthought.com  Fri Nov 16 11:05:20 2007
From: bryanv at enthought.com (Bryan Van de Ven)
Date: Fri, 16 Nov 2007 10:05:20 -0600
Subject: [SciPy-user] adaptive simulated annealing (ASA)
In-Reply-To: <473D749B.6070308@mur.at>
References: <473D749B.6070308@mur.at>
Message-ID: <473DBFC0.4010105@enthought.com>


It's been a while, but I remember ASA having a fairly straightforward interface 
(set up one big config structure, then call one or two functions).  The 'asamin' 
matlab wrapper isn't much code at all, and the guy that wrote that didn't have 
SWIG. :) I doubt you will find many simmulated annealing packages better or more 
featured than ASA, so I'd suggest taking a stab at wrapping it yourself.

http://www.igi.tugraz.at/lehre/MLA/WS01/asamin.html (for reference)
http://www.swig.org/

Georg Holzmann wrote:
> Hallo!
> 
> Is there already a python implementation of the adaptive simulated 
> annealing optimization algorithm (maybe a wrapper to the following C 
> library: http://www.ingber.com/ASA-README.html) ?
> 
> I did quite some google search and did not find something - so any hint 
> is very much appreciated !
> 
> Thanks,
> LG
> Georg
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
> 



From odonnems at yahoo.com  Fri Nov 16 11:24:29 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Fri, 16 Nov 2007 08:24:29 -0800 (PST)
Subject: [SciPy-user] inline.weave compile error
Message-ID: <413269.69593.qm@web58011.mail.re3.yahoo.com>

I am trying to compile some inline c++ code inside python using weave. I always get a similar problem where the compiled file cannot be found (See below output). I am not sure if the problem is with the compiler or something else. I am a new user of scipy and a novice with python so I would appreciate any direction someone can give me because I have not been able to figure out a work around.

Thank you,
Michael

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
I have the following related applications installed:
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Python:
2.4.1 (I am using an older version because this is what ESRI GIS application  supports)
 
Scipy:
scipy-0.6.0.win32-py2.4.exe

MinGW:
MinGW-5.1.3.exe

OS:
Windows XP SP2


+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
When I test the installation of Weave I get the following output:
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>> weave.test()
  Found 1 tests for scipy.weave.ast_tools
  Found 2 tests for scipy.weave.blitz_tools
  Found 9 tests for scipy.weave.build_tools
  Found 0 tests for scipy.weave.c_spec
  Found 26 tests for scipy.weave.catalog
building extensions here: c:\docume~1\michael\locals~1\temp\Michael\python24_compiled\m3
  Found 1 tests for scipy.weave.ext_tools
  Found 0 tests for scipy.weave.inline_tools
  Found 74 tests for scipy.weave.size_check
  Found 16 tests for scipy.weave.slice_handler
  Found 3 tests for scipy.weave.standard_array_spec
  Found 0 tests for __main__
...warning: specified build_dir '_bad_path_' does not exist or is not writable. Trying default locations
.....warning: specified build_dir '_bad_path_' does not exist or is not writable. Trying default locations
............................removing 'c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test' (and everything under it)
error removing c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test: c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test\win3224compiled_catalog: Permission denied
error removing c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test: c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test: Directory not empty
.removing 'c:\docume~1\michael\locals~1\temp\tmpw144aycat_test' (and everything under it)
...............................................................................................
----------------------------------------------------------------------
Ran 132 tests in 2.625s

OK
<unittest.TextTestRunner object at 0x020D6F70>


+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
When I try to test the following script or any other script I get the following message:
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
def prod(m, v):
    #C++ version
    nrows, ncolumns = m.shape

    res = numpy.zeros((nrows, ncolumns), float)
    code = r"""
    for (int i=0; i<nrows; i++)
    {
        for (int j=0; j<ncolumns; j++)
        {
            res(i) += m(i,j)*v(j);
        }
    }
    """

    #err = weave.inline(code,['nrows', 'ncolumns', 'res', 'm', 'v'], type_converters=converters.blitz, compiler='mingw32', verbose=2)  
    err = weave.inline(code,['nrows', 'ncolumns', 'res', 'm', 'v'], verbose=2)

<weave: compiling>
running build_ext
running build_src
building extension "sc_045cfaf40ef1a0738b1066aaa886a55d7" sources
customize Mingw32CCompiler
customize Mingw32CCompiler using build_ext
customize Mingw32CCompiler
customize Mingw32CCompiler using build_ext
building 'sc_045cfaf40ef1a0738b1066aaa886a55d7' extension
compiling C++ sources
C compiler: g++ -mno-cygwin -O2 -Wall

compile options: '-IC:\Python24\lib\site-packages\scipy\weave -IC:\Python24\lib\site-packages\scipy\weave\scxx -IC:\Python24\lib\site-packages\numpy\core\include -IC:\Python24\include -IC:\Python24\PC -c'
g++ -mno-cygwin -O2 -Wall -IC:\Python24\lib\site-packages\scipy\weave -IC:\Python24\lib\site-packages\scipy\weave\scxx -IC:\Python24\lib\site-packages\numpy\core\include -IC:\Python24\include -IC:\Python24\PC -c c:\docume~1\michael\locals~1\temp\Michael\python24_compiled\sc_045cfaf40ef1a0738b1066aaa886a55d7.cpp -o c:\docume~1\michael\locals~1\temp\Michael\python24_intermediate\compiler_c8350d870e6c54e8f29dd7094c2bfb45\Release\docume~1\michael\locals~1\temp\michael\python24_compiled\sc_045cfaf40ef1a0738b1066aaa886a55d7.o
Traceback (most recent call last):
  File "C:\Python24\Lib\site-packages\pythonwin\pywin\framework\scriptutils.py", line 310, in RunScript
    exec codeObject in __main__.__dict__
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 179, in ?
    main()
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 170, in main
    prod(m, v)
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 44, in prod
    err = weave.inline(code,['nrows', 'ncolumns', 'res', 'm', 'v'], verbose=2)
  File "C:\Python24\Lib\site-packages\scipy\weave\inline_tools.py", line 338, in inline
    auto_downcast = auto_downcast,
  File "C:\Python24\Lib\site-packages\scipy\weave\inline_tools.py", line 447, in compile_function
    verbose=verbose, **kw)
  File "C:\Python24\Lib\site-packages\scipy\weave\ext_tools.py", line 365, in compile
    verbose = verbose, **kw)
  File "C:\Python24\Lib\site-packages\scipy\weave\build_tools.py", line 269, in build_extension
    setup(name = module_name, ext_modules = [ext],verbose=verb)
  File "C:\Python24\lib\site-packages\numpy\distutils\core.py", line 173, in setup
    return old_setup(**new_attr)
  File "C:\Python24\lib\distutils\core.py", line 159, in setup
    raise SystemExit, error
CompileError: error: c:\docume~1\michael\locals~1\temp\tmp2dwbkp: No such file or directory





      ____________________________________________________________________________________
Be a better sports nut!  Let your teams follow you 
with Yahoo Mobile. Try it now.  http://mobile.yahoo.com/sports;_ylt=At9_qDKvtAbMuh1G1SQtBI7ntAcJ
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071116/bbcc4897/attachment.html>

From David.L.Goldsmith at noaa.gov  Fri Nov 16 12:43:08 2007
From: David.L.Goldsmith at noaa.gov (David.Goldsmith)
Date: Fri, 16 Nov 2007 09:43:08 -0800
Subject: [SciPy-user] ode on complex systems
In-Reply-To: <web-92145595@uni-stuttgart.de>
References: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
	<1195226348.24691.2.camel@localhost> <web-92145595@uni-stuttgart.de>
Message-ID: <473DD6AC.3070207@noaa.gov>

Nils Wagner wrote:
> AFAIK, you can also apply the patch which is available
> at
>
> http://projects.scipy.org/scipy/scipy/ticket/334
>
>   
> So you can integrate complex ODE's directly without
> doubling the size. Please let me know how it works
> for you.
>   
Please let us all know. :-)

DG

> Nils
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From lbolla at gmail.com  Fri Nov 16 14:08:22 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Fri, 16 Nov 2007 20:08:22 +0100
Subject: [SciPy-user] ode on complex systems
In-Reply-To: <473DD6AC.3070207@noaa.gov>
References: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
	<1195226348.24691.2.camel@localhost> <web-92145595@uni-stuttgart.de>
	<473DD6AC.3070207@noaa.gov>
Message-ID: <80c99e790711161108k6fbdd4b4sd1db90932d1771d6@mail.gmail.com>

I quickly tried to compile scipy with the patch, but it failed (on cygwin).
I had no time to investigate the problem further, but I'll let you know as
soon as possible more details.
L.

PS: for now, I solved the problem by doubling the size of the system as
Fabrice pointed out.

On Nov 16, 2007 6:43 PM, David.Goldsmith <David.L.Goldsmith at noaa.gov> wrote:

> Nils Wagner wrote:
> > AFAIK, you can also apply the patch which is available
> > at
> >
> > http://projects.scipy.org/scipy/scipy/ticket/334
> >
> >
> > So you can integrate complex ODE's directly without
> > doubling the size. Please let me know how it works
> > for you.
> >
> Please let us all know. :-)
>
> DG
>
> > Nils
> >
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071116/4ff633c8/attachment.html>

From pav at iki.fi  Fri Nov 16 15:41:14 2007
From: pav at iki.fi (Pauli Virtanen)
Date: Fri, 16 Nov 2007 20:41:14 +0000 (UTC)
Subject: [SciPy-user] ode on complex systems
References: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
	<1195226348.24691.2.camel@localhost> <web-92145595@uni-stuttgart.de>
	<473DD6AC.3070207@noaa.gov>
	<80c99e790711161108k6fbdd4b4sd1db90932d1771d6@mail.gmail.com>
Message-ID: <fhkv9a$q1g$1@ger.gmane.org>

Fri, 16 Nov 2007 20:08:22 +0100, lorenzo bolla wrote:
> I quickly tried to compile scipy with the patch, but it failed (on
> cygwin). I had no time to investigate the problem further, but I'll let
> you know as soon as possible more details.
> L.

This may be a cygwin-related issue, I'd appreciate more information. The 
patches compile & test fine for me on Ubuntu atop Scipy r3546. (Except 
that the zvode-test.patch needs a rediff.)

-- 
Pauli Virtanen



From yennifersantiago at gmail.com  Fri Nov 16 17:51:22 2007
From: yennifersantiago at gmail.com (Yennifer Santiago)
Date: Fri, 16 Nov 2007 18:51:22 -0400
Subject: [SciPy-user] Biblioteca_algortimos _geneticos
Message-ID: <41bc705b0711161451x15b5a94fvc3f8646bba72cff2@mail.gmail.com>

Hola..

Quisiera saber como se llama la biblioteca de clases de Algoritmos
Geneticos dentro del paquete scipy-0.6.0 que se descarga de internet.
Necesito ver los metodos para operar con Algoritmos Geneticos con que
cuenta esta biblioteca, alli veo varias carpetas pero ninguna
referente a Algoritmos Geneticos.

Gracias

Yennifer Santiago


From cohen at slac.stanford.edu  Fri Nov 16 19:40:40 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Fri, 16 Nov 2007 16:40:40 -0800
Subject: [SciPy-user] Biblioteca_algortimos _geneticos
In-Reply-To: <41bc705b0711161451x15b5a94fvc3f8646bba72cff2@mail.gmail.com>
References: <41bc705b0711161451x15b5a94fvc3f8646bba72cff2@mail.gmail.com>
Message-ID: <473E3888.7010505@slac.stanford.edu>

Yennifer Santiago wrote:
> Hola..
>
> Quisiera saber como se llama la biblioteca de clases de Algoritmos
> Geneticos dentro del paquete scipy-0.6.0 que se descarga de internet.
> Necesito ver los metodos para operar con Algoritmos Geneticos con que
> cuenta esta biblioteca, alli veo varias carpetas pero ninguna
> referente a Algoritmos Geneticos.
>
> Gracias
>
> Yennifer Santiago
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   
se llama ga, dentro sandbox....


From skraelings001 at gmail.com  Fri Nov 16 19:47:38 2007
From: skraelings001 at gmail.com (Reynaldo Baquerizo)
Date: Fri, 16 Nov 2007 19:47:38 -0500
Subject: [SciPy-user] Biblioteca_algortimos _geneticos
In-Reply-To: <41bc705b0711161451x15b5a94fvc3f8646bba72cff2@mail.gmail.com>
References: <41bc705b0711161451x15b5a94fvc3f8646bba72cff2@mail.gmail.com>
Message-ID: <473E3A2A.2020601@gmail.com>

Yennifer Santiago escribi?:
> Hola..
>
> Quisiera saber como se llama la biblioteca de clases de Algoritmos
> Geneticos dentro del paquete scipy-0.6.0 que se descarga de internet.
> Necesito ver los metodos para operar con Algoritmos Geneticos con que
> cuenta esta biblioteca, alli veo varias carpetas pero ninguna
> referente a Algoritmos Geneticos.
>
>   
no tengo la ruta exacta, a?n uso scipy-0.5 pero deber?as ver un 
directorio 'sandbox', dentro de 'sandbox' hay un directorio 'ga' 
(genetic algorithm). Ah? est? todo lo relacionado a algoritmos 
gen?ticos, incluido un programa ejemplo example.py

desde el interprete de python (el standard o ipython) puedes importar el 
m?dulo con ..  'from scipy import ga'
> Gracias
>
> Yennifer Santiago
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   



From aisaac at american.edu  Fri Nov 16 22:54:56 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Fri, 16 Nov 2007 22:54:56 -0500
Subject: [SciPy-user] adaptive simulated annealing (ASA)
In-Reply-To: <473DBFC0.4010105@enthought.com>
References: <473D749B.6070308@mur.at><473DBFC0.4010105@enthought.com>
Message-ID: <Mahogany-0.67.0-1028-20071116-225456.00@american.edu>

On Fri, 16 Nov 2007, Bryan Van de Ven apparently wrote:
> I doubt you will find many simmulated annealing packages better or more 
> featured than ASA, so I'd suggest taking a stab at wrapping it yourself. 
> http://www.igi.tugraz.at/lehre/MLA/WS01/asamin.html (for reference)
> http://www.swig.org/ 


Please keep the list apprised of your efforts.

Thank you,
Alan Isaac






From dwarnold45 at suddenlink.net  Sat Nov 17 02:38:49 2007
From: dwarnold45 at suddenlink.net (David Arnold)
Date: Fri, 16 Nov 2007 23:38:49 -0800
Subject: [SciPy-user] Instruction in math and science
Message-ID: <C238A75E-591A-4B42-B9E7-3A2068DB4206@suddenlink.net>

All,

We are considering using Python and some of its libraries to aid  
instruction in mathematics and science at our college. I've read much  
of the documentation and installation directions for python and some  
of its libraries, such as matplotlib, numpy, and scipy.

I am aware that there are binary installations for the Mac at:

http://www.pythonmac.org/packages/

But no binary for scipy at this site for the 2.5x version of python.

I am also aware of the enthought distribution for windows and the  
installer at:

http://code.enthought.com/downloads/enthon-python2.4-1.0.0.exe

I am looking for advice from experienced educators who have advised  
their students on installation on Mac and Windows. These students at  
our school are unlikely to have the expertise to unzip, configure,  
compile, etc., and need a "one-click" solution to install the  
software needed.

Can anyone share their experience with their students and/or give  
advice?

Thanks.

David Arnold
College of the Redwoods
http://online.redwoods.edu/instruct/darnold/


From wdj at usna.edu  Sat Nov 17 07:22:41 2007
From: wdj at usna.edu (David Joyner)
Date: Sat, 17 Nov 2007 07:22:41 -0500 (EST)
Subject: [SciPy-user] Instruction in math and science
Message-ID: <20071117072241.AJK54693@mp2.nettest.usna.edu>

David:
I'd like to suggest Sage, 
http://www.sagemath.org/
which comes with python 2.5 and scipy
http://www.sagemath.org/packages/standard/
and is available in binary for macs and windows
http://www.sagemath.org/download.html
I've been using it for teaching DEs
http://sage.math.washington.edu/home/wdj/teaching/index.html
http://cadigweb.ew.usna.edu/%7Ewdj/teach/sm212/
- David Joyner

---- Original message ----
>Date: Fri, 16 Nov 2007 23:38:49 -0800
>From: David Arnold <dwarnold45 at suddenlink.net>  
>Subject: [SciPy-user] Instruction in math and science  
>To: scipy-user at scipy.org
>
>All,
>
>We are considering using Python and some of its libraries to aid  
>instruction in mathematics and science at our college. I've read much  
>of the documentation and installation directions for python and some  
>of its libraries, such as matplotlib, numpy, and scipy.
>
>I am aware that there are binary installations for the Mac at:
>
>http://www.pythonmac.org/packages/
>
>But no binary for scipy at this site for the 2.5x version of python.
>
>I am also aware of the enthought distribution for windows and the  
>installer at:
>
>http://code.enthought.com/downloads/enthon-python2.4-1.0.0.exe
>
>I am looking for advice from experienced educators who have advised  
>their students on installation on Mac and Windows. These students at  
>our school are unlikely to have the expertise to unzip, configure,  
>compile, etc., and need a "one-click" solution to install the  
>software needed.
>
>Can anyone share their experience with their students and/or give  
>advice?
>
>Thanks.
>
>David Arnold
>College of the Redwoods
>http://online.redwoods.edu/instruct/darnold/
>_______________________________________________
>SciPy-user mailing list
>SciPy-user at scipy.org
>http://projects.scipy.org/mailman/listinfo/scipy-user


From boris at vonloesch.de  Sat Nov 17 12:05:05 2007
From: boris at vonloesch.de (Boris von Loesch)
Date: Sat, 17 Nov 2007 18:05:05 +0100
Subject: [SciPy-user] Problems compiling Scipy with Cygwin
Message-ID: <41a2e2dc0711170905j635a61ofc6b8e6727e3299e@mail.gmail.com>

Hello,

I tried to compile scipy with cygwin, but the building ended in an
error. I use the latest Atlas 3.8.0, and numpy 1.0.4 compiled without
problems using

python.exe setup.py config --compiler=mingw32 build --compiler=mingw32
bdist_wininst

If I build scipy the following error occur:

...
building 'scipy.integrate.vode' extension
compiling C sources
C compiler: gcc -mno-cygwin -O2 -Wall -Wstrict-prototypes

compile options: '-DATLAS_INFO="\"?.?.?\"" -Ic:\cygwin\usr\include -Ibuild\src.w
in32-2.5 -Ic:\programme\python2.5\lib\site-packages\numpy\core\include -Ic:\prog
ramme\python2.5\include -Ic:\programme\python2.5\PC -c'
C:\cygwin\bin\g77.exe -g -Wall -mno-cygwin -g -Wall -mno-cygwin -shared build\te
mp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o build\temp
.win32-2.5\Release\build\src.win32-2.5\fortranobject.o -Lc:\cygwin\usr\lib -Lc:\
programme\python2.5\libs -Lc:\programme\python2.5\PCBuild -Lbuild\temp.win32-2.5
 -lodepack -llinpack_lite -lmach -lf77blas -lcblas -latlas -lpython25 -lg2c -o b
uild\lib.win32-2.5\scipy\integrate\vode.pyd
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0x367): undefined reference to `__impure_ptr'
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0x462): undefined reference to `__impure_ptr'
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0x585): undefined reference to `__impure_ptr'
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0x9b1): undefined reference to `__impure_ptr'
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0xb23): undefined reference to `__impure_ptr'
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0xb79): more undefined references to `__impure_ptr' follow
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0x18f4): undefined reference to `_setjmp'
build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemodule.o:vo
demodule.c:(.text+0x1ac5): undefined reference to `_setjmp'
collect2: ld returned 1 exit status
error: Command "C:\cygwin\bin\g77.exe -g -Wall -mno-cygwin -g -Wall -mno-cygwin
-shared build\temp.win32-2.5\Release\build\src.win32-2.5\scipy\integrate\vodemod
ule.o build\temp.win32-2.5\Release\build\src.win32-2.5\fortranobject.o -Lc:\cygw
in\usr\lib -Lc:\programme\python2.5\libs -Lc:\programme\python2.5\PCBuild -Lbuil
d\temp.win32-2.5 -lodepack -llinpack_lite -lmach -lf77blas -lcblas -latlas -lpyt
hon25 -lg2c -o build\lib.win32-2.5\scipy\integrate\vode.pyd" failed with exit st
atus 1

I also tried to build scipy with "-fno-second-underscore" enabled via
F77FLAGS, but that doesn't change anything.

Does anybody know how to fix this?

Best Regards,
Boris


From alan at ajackson.org  Sat Nov 17 16:30:10 2007
From: alan at ajackson.org (Alan Jackson)
Date: Sat, 17 Nov 2007 15:30:10 -0600
Subject: [SciPy-user] 2D arrays in weave.inline
Message-ID: <20071117153010.6019d1df@nova.oplnk.net>

This is probably an astonishingly simple question, but I've been struggling for
some time with it.

I am trying to work with weave for the first time, and it is becoming clear that
I don't understand how 2D arrays get passed. The example below illustrates the problem I'm
having... the 2D indexing doesn't seem to be working right at all.

BTW the link to the weave documentation is a dead link. Is there any documentation beyond
a couple of short examples? Every promising google link turns up dead.


#!/usr/bin/env python

from numpy import array, zeros, abs, ones, arange, ravel, nan, empty, isnan, histogram, hstack, multiply, log, shape, triu
import scipy.weave as weave
from scipy.weave import converters

def Matrix(width, minval, maxval, exponent, type="HorseSaddle"):

    halfSqrt2 = 0.707106781185
    sin45 = halfSqrt2
    cos45 = halfSqrt2
    Matrix = zeros((width, width), dtype=float)
    for j in range(0,width):
        jScaled = j / (halfSqrt2 * (width-1))
        jTranslated = jScaled - halfSqrt2
        jsin45 = jTranslated * sin45
        jcos45 = jsin45
        for i in range(j,width):
          iScaled = i / (halfSqrt2 * (width-1));
          iTranslated = iScaled - halfSqrt2;
          iRotated = iTranslated * cos45 + jsin45;
          jRotated = - iTranslated * sin45 + jcos45;
          iRotated = abs(iRotated);
          jRotated = abs(jRotated);
          Matrix[i, j] = round(maxval * iRotated**exponent + \
                                      minval * jRotated**exponent, 2)
          Matrix[j,i] = Matrix[i,j]

    return Matrix

################################################################################
#   algorithm - WEAVE version
################################################################################

def Fast(s1, s2, wt1, wt2, M):
    
    seq1 = s1[0]
    seq2 = s2[0]
    t1 = s1[1]
    t2 = s2[1]
    
    code = '''

  
  for (unsigned int idx2 = 0; idx2 < 10; ++idx2) {
    
    printf("\\n");
    for (unsigned int idx1 = 0; idx1 < 10; ++idx1) {
      printf("%6f  ", M[idx1, idx2]);
      
    }// for whole s1
  }// for whole s2
  printf("\\n");
  '''

    weave.inline(code, ["seq1", "seq2", "t1", "t2", "M" ])

################################################################################
#		test section
################################################################################

if __name__ == '__main__' :
    attr1  = array([1,2,2,3,2,4,3,5,4,3,3,  3,4,6,1,0,2,4,2,5,2,3])
    attr2 = array([1,2,2,3,2,5,3,5,4,3,2,6, 3,4,6,1,0,1,4,2,5,2,3])
    t = arange(1, len(attr1)+1, dtype=float)
    wt1 = ones(len(attr1), dtype=float)
    t2 = arange(1, len(attr2)+1, dtype=float)
    wt2 = ones(len(attr2), dtype=float)

    seq1 = [attr1, t]
    seq2 = [attr2, t2]
    M = Matrix(7,-10,10,2.)
    print M
    s1s2, s2s1 = Fast(seq1, seq2, wt1, wt2, M )


-- 
-----------------------------------------------------------------------
| Alan K. Jackson            | To see a World in a Grain of Sand      |
| alan at ajackson.org          | And a Heaven in a Wild Flower,         |
| www.ajackson.org           | Hold Infinity in the palm of your hand |
| Houston, Texas             | And Eternity in an hour. - Blake       |
-----------------------------------------------------------------------


From gael.varoquaux at normalesup.org  Sat Nov 17 16:46:59 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sat, 17 Nov 2007 22:46:59 +0100
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117153010.6019d1df@nova.oplnk.net>
References: <20071117153010.6019d1df@nova.oplnk.net>
Message-ID: <20071117214659.GE12125@clipper.ens.fr>

On Sat, Nov 17, 2007 at 03:30:10PM -0600, Alan Jackson wrote:
> This is probably an astonishingly simple question, but I've been
> struggling for some time with it.

> I am trying to work with weave for the first time, and it is becoming
> clear that I don't understand how 2D arrays get passed.

Me neither. This is why I use the blitz converter. I modified your
example to use the vlitz converters, it is just so much nicer:

++++++++++++++++++++++++++++++++++++++++++++
def Fast(s1, s2, wt1, wt2, M):

    seq1 = s1[0]
    seq2 = s2[0]
    t1 = s1[1]
    t2 = s2[1]

    code = '''


  for (int idx2 = 0; idx2 < 10; ++idx2) {
    printf("\\n");
    for (int idx1 = 0; idx1 < 10; ++idx1) {
      printf("%6f  ", M(idx1, idx2));

    }// for whole s1
  }// for whole s2
  printf("\\n");
  '''

    weave.inline(code, ["seq1", "seq2", "t1", "t2", "M" ],
                        type_converters=converters.blitz,
            )
++++++++++++++++++++++++++++++++++++++++++++


> BTW the link to the weave documentation is a dead link. Is there any
> documentation beyond a couple of short examples? Every promising google
> link turns up dead.

There have been a few threads on this mailing list. I also think
http://scipy.org/PerformancePython#head-a3f4dd816378d3ba4cbdd3d23dc98529e8ad7087
can be useful.

Ga?l


From william.ratcliff at gmail.com  Sat Nov 17 17:16:55 2007
From: william.ratcliff at gmail.com (william ratcliff)
Date: Sat, 17 Nov 2007 17:16:55 -0500
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117214659.GE12125@clipper.ens.fr>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
Message-ID: <827183970711171416v35880e81h1d377b155fca1bde@mail.gmail.com>

Gael, have you tried weave with openmp?

Cheers,
William

On Nov 17, 2007 4:46 PM, Gael Varoquaux <gael.varoquaux at normalesup.org> wrote:
> On Sat, Nov 17, 2007 at 03:30:10PM -0600, Alan Jackson wrote:
> > This is probably an astonishingly simple question, but I've been
> > struggling for some time with it.
>
> > I am trying to work with weave for the first time, and it is becoming
> > clear that I don't understand how 2D arrays get passed.
>
> Me neither. This is why I use the blitz converter. I modified your
> example to use the vlitz converters, it is just so much nicer:
>
> ++++++++++++++++++++++++++++++++++++++++++++
> def Fast(s1, s2, wt1, wt2, M):
>
>     seq1 = s1[0]
>     seq2 = s2[0]
>     t1 = s1[1]
>     t2 = s2[1]
>
>     code = '''
>
>
>   for (int idx2 = 0; idx2 < 10; ++idx2) {
>     printf("\\n");
>     for (int idx1 = 0; idx1 < 10; ++idx1) {
>       printf("%6f  ", M(idx1, idx2));
>
>     }// for whole s1
>   }// for whole s2
>   printf("\\n");
>   '''
>
>     weave.inline(code, ["seq1", "seq2", "t1", "t2", "M" ],
>                         type_converters=converters.blitz,
>             )
> ++++++++++++++++++++++++++++++++++++++++++++
>
>
> > BTW the link to the weave documentation is a dead link. Is there any
> > documentation beyond a couple of short examples? Every promising google
> > link turns up dead.
>
> There have been a few threads on this mailing list. I also think
> http://scipy.org/PerformancePython#head-a3f4dd816378d3ba4cbdd3d23dc98529e8ad7087
> can be useful.
>
> Ga?l
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From gael.varoquaux at normalesup.org  Sat Nov 17 17:20:26 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sat, 17 Nov 2007 23:20:26 +0100
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <827183970711171416v35880e81h1d377b155fca1bde@mail.gmail.com>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<827183970711171416v35880e81h1d377b155fca1bde@mail.gmail.com>
Message-ID: <20071117222026.GG12125@clipper.ens.fr>

On Sat, Nov 17, 2007 at 05:16:55PM -0500, william ratcliff wrote:
> Gael, have you tried weave with openmp?

No, and I would be interested to find out if it works or not.

But if it doesn't, I don't see why ctypes would not (maybe I am missing
some thing, though, my knowledge of dynamical loading is very thin).

Ga?l


From alan at ajackson.org  Sat Nov 17 17:44:52 2007
From: alan at ajackson.org (Alan Jackson)
Date: Sat, 17 Nov 2007 16:44:52 -0600
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117214659.GE12125@clipper.ens.fr>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
Message-ID: <20071117164452.4d0ca4c7@nova.oplnk.net>

On Sat, 17 Nov 2007 22:46:59 +0100
Gael Varoquaux <gael.varoquaux at normalesup.org> wrote:

> On Sat, Nov 17, 2007 at 03:30:10PM -0600, Alan Jackson wrote:
> > This is probably an astonishingly simple question, but I've been
> > struggling for some time with it.
> 
> > I am trying to work with weave for the first time, and it is becoming
> > clear that I don't understand how 2D arrays get passed.
> 
> Me neither. This is why I use the blitz converter. I modified your
> example to use the vlitz converters, it is just so much nicer:
> 

Well, with my *real* code, blitz gave me several pages of errors - I tried
treating the 2D arrays as a 1D array, doing all the index arithmetic myself,
and that seems to work. Which strikes me as odd...

-- 
-----------------------------------------------------------------------
| Alan K. Jackson            | To see a World in a Grain of Sand      |
| alan at ajackson.org          | And a Heaven in a Wild Flower,         |
| www.ajackson.org           | Hold Infinity in the palm of your hand |
| Houston, Texas             | And Eternity in an hour. - Blake       |
-----------------------------------------------------------------------


From stefan at sun.ac.za  Sat Nov 17 18:13:44 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Sun, 18 Nov 2007 01:13:44 +0200
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117164452.4d0ca4c7@nova.oplnk.net>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<20071117164452.4d0ca4c7@nova.oplnk.net>
Message-ID: <20071117231344.GC26080@mentat.za.net>

On Sat, Nov 17, 2007 at 04:44:52PM -0600, Alan Jackson wrote:
> On Sat, 17 Nov 2007 22:46:59 +0100
> Well, with my *real* code, blitz gave me several pages of errors - I tried
> treating the 2D arrays as a 1D array, doing all the index arithmetic myself,
> and that seems to work. Which strikes me as odd...

With Blitz, you are allowed M(i,j) indexing.  Otherwise, only the
memory address of the data is passed and, as you noticed, you must
calculate the offset yourself, i.e. M[i*nr_cols + j].

I modified your code to work with ctypes, which you may find
interesting.

Use

  python setup.py build_ext -i

to build the C file.  Then run alan.py.

Cheers
St?fan
-------------- next part --------------
A non-text attachment was scrubbed...
Name: setup.py
Type: text/x-python
Size: 461 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071118/bf802271/attachment.py>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: alan.c
Type: text/x-csrc
Size: 243 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071118/bf802271/attachment.c>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: alan.py
Type: text/x-python
Size: 1973 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071118/bf802271/attachment-0001.py>

From william.ratcliff at gmail.com  Sat Nov 17 18:16:26 2007
From: william.ratcliff at gmail.com (william ratcliff)
Date: Sat, 17 Nov 2007 18:16:26 -0500
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117164452.4d0ca4c7@nova.oplnk.net>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<20071117164452.4d0ca4c7@nova.oplnk.net>
Message-ID: <827183970711171516n5d439908w55bb68c70abcb4f7@mail.gmail.com>

I had good luck with it using blitz type converters (though if there's
a bug in the C code, it's painful to find ;>).  Did you look at the
array3d example that came with the code?  So far, I've been able to
pass 1,2, and 3d arrays successfully.  But, I've been cheating and
creating the arrays that I want returned in numpy before calling the
routine in weave and passing them in as arguments.  I modify them
within weave, and then check the results when weave terminates to make
sure they have the values I think they should have in numpy.  Then I
didn't have to work on figuring out return types, allocating
memory,etc.  I also use blitz type converters.

What kind of errors are you getting?  When I was writing my first
program using weave, I got tons of errors--even if it was something as
minor as forgetting a ";" in one of the lines of code--one helpful
thing to do is to find the temporary directory where weave stores the
c++ versions of your code and deleting it--there should be two
directories.  However, if it's not a simple question of stale code
lying around, then wade through the pages of output and try to find
out whether there is a simple bug in your C code.  The next possible
source of error could be a compiler issue--for example, I normally use
an older version of gcc (I forget, something like 3.4) and upgraded to
version 4.x (one of the most recent) and found that it no longer
worked.  Is the real code involved something simple enough that you
can post it?

Cheers,
William

On Nov 17, 2007 5:44 PM, Alan Jackson <alan at ajackson.org> wrote:
> On Sat, 17 Nov 2007 22:46:59 +0100
> Gael Varoquaux <gael.varoquaux at normalesup.org> wrote:
>
> > On Sat, Nov 17, 2007 at 03:30:10PM -0600, Alan Jackson wrote:
> > > This is probably an astonishingly simple question, but I've been
> > > struggling for some time with it.
> >
> > > I am trying to work with weave for the first time, and it is becoming
> > > clear that I don't understand how 2D arrays get passed.
> >
> > Me neither. This is why I use the blitz converter. I modified your
> > example to use the vlitz converters, it is just so much nicer:
> >
>
> Well, with my *real* code, blitz gave me several pages of errors - I tried
> treating the 2D arrays as a 1D array, doing all the index arithmetic myself,
> and that seems to work. Which strikes me as odd...
>
> --
> -----------------------------------------------------------------------
> | Alan K. Jackson            | To see a World in a Grain of Sand      |
> | alan at ajackson.org          | And a Heaven in a Wild Flower,         |
> | www.ajackson.org           | Hold Infinity in the palm of your hand |
> | Houston, Texas             | And Eternity in an hour. - Blake       |
> -----------------------------------------------------------------------
> _______________________________________________
>
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From alan at ajackson.org  Sat Nov 17 18:30:18 2007
From: alan at ajackson.org (Alan Jackson)
Date: Sat, 17 Nov 2007 17:30:18 -0600
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <827183970711171516n5d439908w55bb68c70abcb4f7@mail.gmail.com>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<20071117164452.4d0ca4c7@nova.oplnk.net>
	<827183970711171516n5d439908w55bb68c70abcb4f7@mail.gmail.com>
Message-ID: <20071117173018.582e94a7@nova.oplnk.net>

Thanks all for your help. The path I chose was to redo my indexing - that was
the path of least resistance. I'm not sure where the blitz errors came from -
the same code compiled clean with inline, so that is a mystery (I did modify
the arrays for blitz before compiling).

I also created all the arrays I wanted returned in numpy first - which makes for
a really long function call, but it works and it is not too awkward.

I got a speed up of x37 on my real test. 

On Sat, 17 Nov 2007 18:16:26 -0500
"william ratcliff" <william.ratcliff at gmail.com> wrote:

> I had good luck with it using blitz type converters (though if there's
> a bug in the C code, it's painful to find ;>).  Did you look at the
> array3d example that came with the code?  So far, I've been able to
> pass 1,2, and 3d arrays successfully.  But, I've been cheating and
> creating the arrays that I want returned in numpy before calling the
> routine in weave and passing them in as arguments.  I modify them
> within weave, and then check the results when weave terminates to make
> sure they have the values I think they should have in numpy.  Then I
> didn't have to work on figuring out return types, allocating
> memory,etc.  I also use blitz type converters.
> 
> What kind of errors are you getting?  When I was writing my first
> program using weave, I got tons of errors--even if it was something as
> minor as forgetting a ";" in one of the lines of code--one helpful
> thing to do is to find the temporary directory where weave stores the
> c++ versions of your code and deleting it--there should be two
> directories.  However, if it's not a simple question of stale code
> lying around, then wade through the pages of output and try to find
> out whether there is a simple bug in your C code.  The next possible
> source of error could be a compiler issue--for example, I normally use
> an older version of gcc (I forget, something like 3.4) and upgraded to
> version 4.x (one of the most recent) and found that it no longer
> worked.  Is the real code involved something simple enough that you
> can post it?
> 
> Cheers,
> William
> 
> On Nov 17, 2007 5:44 PM, Alan Jackson <alan at ajackson.org> wrote:
> > On Sat, 17 Nov 2007 22:46:59 +0100
> > Gael Varoquaux <gael.varoquaux at normalesup.org> wrote:
> >
> > > On Sat, Nov 17, 2007 at 03:30:10PM -0600, Alan Jackson wrote:
> > > > This is probably an astonishingly simple question, but I've been
> > > > struggling for some time with it.
> > >
> > > > I am trying to work with weave for the first time, and it is becoming
> > > > clear that I don't understand how 2D arrays get passed.
> > >
> > > Me neither. This is why I use the blitz converter. I modified your
> > > example to use the vlitz converters, it is just so much nicer:
> > >
> >
> > Well, with my *real* code, blitz gave me several pages of errors - I tried
> > treating the 2D arrays as a 1D array, doing all the index arithmetic myself,
> > and that seems to work. Which strikes me as odd...
> >
> > --
> > -----------------------------------------------------------------------
> > | Alan K. Jackson            | To see a World in a Grain of Sand      |
> > | alan at ajackson.org          | And a Heaven in a Wild Flower,         |
> > | www.ajackson.org           | Hold Infinity in the palm of your hand |
> > | Houston, Texas             | And Eternity in an hour. - Blake       |
> > -----------------------------------------------------------------------
> > _______________________________________________
> >
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user


-- 
-----------------------------------------------------------------------
| Alan K. Jackson            | To see a World in a Grain of Sand      |
| alan at ajackson.org          | And a Heaven in a Wild Flower,         |
| www.ajackson.org           | Hold Infinity in the palm of your hand |
| Houston, Texas             | And Eternity in an hour. - Blake       |
-----------------------------------------------------------------------


From gael.varoquaux at normalesup.org  Sat Nov 17 18:33:06 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 18 Nov 2007 00:33:06 +0100
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117164452.4d0ca4c7@nova.oplnk.net>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<20071117164452.4d0ca4c7@nova.oplnk.net>
Message-ID: <20071117233306.GH12125@clipper.ens.fr>

On Sat, Nov 17, 2007 at 04:44:52PM -0600, Alan Jackson wrote:
> Well, with my *real* code, blitz gave me several pages of errors 

If you are doing complicated things, you should really use another
technique to call C from Python. Ctypes is pretty easy to use
(http://www.scipy.org/Cookbook/Ctypes). Debug is also much easier,
because it is easier to understand the compiler's error.

Ga?l


From gael.varoquaux at normalesup.org  Sat Nov 17 18:36:06 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 18 Nov 2007 00:36:06 +0100
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117173018.582e94a7@nova.oplnk.net>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<20071117164452.4d0ca4c7@nova.oplnk.net>
	<827183970711171516n5d439908w55bb68c70abcb4f7@mail.gmail.com>
	<20071117173018.582e94a7@nova.oplnk.net>
Message-ID: <20071117233606.GI12125@clipper.ens.fr>

On Sat, Nov 17, 2007 at 05:30:18PM -0600, Alan Jackson wrote:
> I'm not sure where the blitz errors came from - the same code compiled
> clean with inline, so that is a mystery (I did modify the arrays for
> blitz before compiling).

Just checking: you did replace the "M[x, y]" by "M(x, y)" in all your
code? 

> I also created all the arrays I wanted returned in numpy first - which
> makes for a really long function call, but it works and it is not too
> awkward.

I always do that. It removes all memory management from the C level.
Tht's one less hurdle. I usual wrap my function in a another one, to hide
this ugly function call, though.

> I got a speed up of x37 on my real test. 

Nice !

Ga?l


From prabhu at aero.iitb.ac.in  Sat Nov 17 20:13:43 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Sun, 18 Nov 2007 06:43:43 +0530
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117231344.GC26080@mentat.za.net>
References: <20071117153010.6019d1df@nova.oplnk.net>	<20071117214659.GE12125@clipper.ens.fr>	<20071117164452.4d0ca4c7@nova.oplnk.net>
	<20071117231344.GC26080@mentat.za.net>
Message-ID: <473F91C7.3030202@aero.iitb.ac.in>

Stefan van der Walt wrote:
> On Sat, Nov 17, 2007 at 04:44:52PM -0600, Alan Jackson wrote:
>> On Sat, 17 Nov 2007 22:46:59 +0100
>> Well, with my *real* code, blitz gave me several pages of errors - I tried
>> treating the 2D arrays as a 1D array, doing all the index arithmetic myself,
>> and that seems to work. Which strikes me as odd...
> 
> With Blitz, you are allowed M(i,j) indexing.  Otherwise, only the
> memory address of the data is passed and, as you noticed, you must
> calculate the offset yourself, i.e. M[i*nr_cols + j].

Travis added a few macros to automate this calculation (for non-blitz 
cases).  If you passed in 'a' then A1(i), A2(i, j) A3(i, j, k) and A4(i, 
j, k, l) are all defined.  Here are the macros:

#define A1(i) (*((long*)(a_array->data + (i)*Sa[0])))
#define A2(i,j) (*((long*)(a_array->data + (i)*Sa[0] + (j)*Sa[1])))
#define A3(i,j,k) (*((long*)(a_array->data + (i)*Sa[0] + (j)*Sa[1] + 
(k)*Sa[2])))
#define A4(i,j,k,l) (*((long*)(a_array->data + (i)*Sa[0] + (j)*Sa[1] + 
(k)*Sa[2] + (l)*Sa[3])))


cheers,
prabhu


From alan at ajackson.org  Sat Nov 17 22:16:34 2007
From: alan at ajackson.org (Alan Jackson)
Date: Sat, 17 Nov 2007 21:16:34 -0600
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117233606.GI12125@clipper.ens.fr>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<20071117164452.4d0ca4c7@nova.oplnk.net>
	<827183970711171516n5d439908w55bb68c70abcb4f7@mail.gmail.com>
	<20071117173018.582e94a7@nova.oplnk.net>
	<20071117233606.GI12125@clipper.ens.fr>
Message-ID: <20071117211634.7cd27b9a@nova.oplnk.net>

On Sun, 18 Nov 2007 00:36:06 +0100
Gael Varoquaux <gael.varoquaux at normalesup.org> wrote:

> On Sat, Nov 17, 2007 at 05:30:18PM -0600, Alan Jackson wrote:
> > I'm not sure where the blitz errors came from - the same code compiled
> > clean with inline, so that is a mystery (I did modify the arrays for
> > blitz before compiling).
> 
> Just checking: you did replace the "M[x, y]" by "M(x, y)" in all your
> code?
 
Yep...


> 
> > I also created all the arrays I wanted returned in numpy first - which
> > makes for a really long function call, but it works and it is not too
> > awkward.
> 
> I always do that. It removes all memory management from the C level.
> Tht's one less hurdle. I usual wrap my function in a another one, to hide
> this ugly function call, though.
> 
> > I got a speed up of x37 on my real test. 
> 
> Nice !
> 
> Ga?l
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user


-- 
-----------------------------------------------------------------------
| Alan K. Jackson            | To see a World in a Grain of Sand      |
| alan at ajackson.org          | And a Heaven in a Wild Flower,         |
| www.ajackson.org           | Hold Infinity in the palm of your hand |
| Houston, Texas             | And Eternity in an hour. - Blake       |
-----------------------------------------------------------------------


From alan at ajackson.org  Sat Nov 17 22:17:39 2007
From: alan at ajackson.org (Alan Jackson)
Date: Sat, 17 Nov 2007 21:17:39 -0600
Subject: [SciPy-user] 2D arrays in weave.inline
In-Reply-To: <20071117233306.GH12125@clipper.ens.fr>
References: <20071117153010.6019d1df@nova.oplnk.net>
	<20071117214659.GE12125@clipper.ens.fr>
	<20071117164452.4d0ca4c7@nova.oplnk.net>
	<20071117233306.GH12125@clipper.ens.fr>
Message-ID: <20071117211739.00075255@nova.oplnk.net>

I'll have to look into it...

On Sun, 18 Nov 2007 00:33:06 +0100
Gael Varoquaux <gael.varoquaux at normalesup.org> wrote:

> On Sat, Nov 17, 2007 at 04:44:52PM -0600, Alan Jackson wrote:
> > Well, with my *real* code, blitz gave me several pages of errors 
> 
> If you are doing complicated things, you should really use another
> technique to call C from Python. Ctypes is pretty easy to use
> (http://www.scipy.org/Cookbook/Ctypes). Debug is also much easier,
> because it is easier to understand the compiler's error.
> 
> Ga?l
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user


-- 
-----------------------------------------------------------------------
| Alan K. Jackson            | To see a World in a Grain of Sand      |
| alan at ajackson.org          | And a Heaven in a Wild Flower,         |
| www.ajackson.org           | Hold Infinity in the palm of your hand |
| Houston, Texas             | And Eternity in an hour. - Blake       |
-----------------------------------------------------------------------


From dwarnold45 at suddenlink.net  Sun Nov 18 00:06:10 2007
From: dwarnold45 at suddenlink.net (David Arnold)
Date: Sat, 17 Nov 2007 21:06:10 -0800
Subject: [SciPy-user] Installing numpy
Message-ID: <E1EE2D9C-7CA0-44A2-9094-DF743AE8EF2F@suddenlink.net>

All,

Mac OS X 10.4.11 PPC.

Installed python from this:

http://www.python.org/ftp/python/2.5.1/python-2.5.1-macosx.dmg

Installation finishes with "There were errors installing the  
software. Please try installing again." Installing again makes no  
difference, but Python appears to be running:

darnold $ python
Python 2.5.1 (r251:54869, Apr 18 2007, 22:08:04)
[GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>>

Downloaded numpy from:  http://downloads.sourceforge.net/numpy/ 
numpy-1.0.4.tar.gz?use_mirror=internap

After unpacking and changing directories, simply did this:

numpy-1.0.4 $ python setup.py install

Seem to finish OK. But testing according to README.txt resulted in:

numpy-1.0.4 $ python -c 'import numpy; numpy.test()'
Running from numpy source directory.
Traceback (most recent call last):
   File "<string>", line 1, in <module>
AttributeError: 'module' object has no attribute 'test'

Any ideas or advice?

Thanks

David.






From dwarnold45 at suddenlink.net  Sun Nov 18 00:27:14 2007
From: dwarnold45 at suddenlink.net (David Arnold)
Date: Sat, 17 Nov 2007 21:27:14 -0800
Subject: [SciPy-user] Installing Scipy
Message-ID: <4F98C7F5-CA90-4568-839D-0CD8D9295A4A@suddenlink.net>

All,

Mac OS X 10.4.11 PPC.

Installed python from this:

http://www.python.org/ftp/python/2.5.1/python-2.5.1-macosx.dmg

Installation finishes with "There were errors installing the
software. Please try installing again." Installing again makes no
difference, but Python appears to be running:

scipy-0.6.0 $ python
Python 2.5.1 (r251:54869, Apr 18 2007, 22:08:04)
[GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>>


Downloaded scipy from: http://prdownloads.sourceforge.net/scipy/ 
scipy-0.6.0.tar.gz?download

After unpacking and changing directories, simply did this:

scipy-0.6.0 $ sudo python setup.py install

Seem to finish OK. Lots of "Warnings.' But testing according to  
INSTALL.txt resulted in:

scipy-0.6.0 $ python
Python 2.5.1 (r251:54869, Apr 18 2007, 22:08:04)
[GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import scipy
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
   File "scipy/__init__.py", line 54, in <module>
     from __config__ import show as show_config
ImportError: No module named __config__


Any ideas or advice?

Thanks

David.



From robert.kern at gmail.com  Sun Nov 18 00:41:59 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Sat, 17 Nov 2007 23:41:59 -0600
Subject: [SciPy-user] Installing numpy
In-Reply-To: <E1EE2D9C-7CA0-44A2-9094-DF743AE8EF2F@suddenlink.net>
References: <E1EE2D9C-7CA0-44A2-9094-DF743AE8EF2F@suddenlink.net>
Message-ID: <473FD0A7.3030404@gmail.com>

David Arnold wrote:
> All,
> 
> Mac OS X 10.4.11 PPC.
> 
> Installed python from this:
> 
> http://www.python.org/ftp/python/2.5.1/python-2.5.1-macosx.dmg
> 
> Installation finishes with "There were errors installing the  
> software. Please try installing again." Installing again makes no  
> difference, but Python appears to be running:
> 
> darnold $ python
> Python 2.5.1 (r251:54869, Apr 18 2007, 22:08:04)
> [GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin
> Type "help", "copyright", "credits" or "license" for more information.
>  >>>
> 
> Downloaded numpy from:  http://downloads.sourceforge.net/numpy/ 
> numpy-1.0.4.tar.gz?use_mirror=internap
> 
> After unpacking and changing directories, simply did this:
> 
> numpy-1.0.4 $ python setup.py install
> 
> Seem to finish OK. But testing according to README.txt resulted in:
> 
> numpy-1.0.4 $ python -c 'import numpy; numpy.test()'
> Running from numpy source directory.
> Traceback (most recent call last):
>    File "<string>", line 1, in <module>
> AttributeError: 'module' object has no attribute 'test'
> 
> Any ideas or advice?

Change to another directory. The numpy/ directory in the source tree is getting
picked up instead of the installed one. Same thing with scipy.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From dwarnold45 at suddenlink.net  Sun Nov 18 01:02:39 2007
From: dwarnold45 at suddenlink.net (David Arnold)
Date: Sat, 17 Nov 2007 22:02:39 -0800
Subject: [SciPy-user] Installing numpy
In-Reply-To: <473FD0A7.3030404@gmail.com>
References: <E1EE2D9C-7CA0-44A2-9094-DF743AE8EF2F@suddenlink.net>
	<473FD0A7.3030404@gmail.com>
Message-ID: <7F5B3638-3AF9-4181-9A52-7908DAB7A425@suddenlink.net>

Wow. Good catch. Switching directories:

Ran 692 tests in 1.211s

 >>>import numpy
 >>>numpy.test()

...

OK
<unittest._TextTestResult run=692 errors=0 failures=0>
 >>>


Not so lucky with scipy:

darnold $ python
Python 2.5.1 (r251:54869, Apr 18 2007, 22:08:04)
[GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import scipy
s>>> scipy.test(level=1)

....

Ran 1716 tests in 6.949s

FAILED (failures=2)
<unittest._TextTestResult run=1716 errors=0 failures=2>
 >>>


Along the way, I got:

****************************************************************
WARNING: clapack module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
   then scipy uses flapack instead of clapack.
****************************************************************

And:

Warning: FAILURE importing tests for <module 'scipy.linsolve.umfpack'  
from '.../linsolve/umfpack/__init__.pyc'>
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site- 
packages/scipy/linsolve/umfpack/tests/test_umfpack.py:17:  
AttributeError: 'module' object has no attribute 'umfpack' (in <module>)
Warning: FAILURE importing tests for <module  
'scipy.linsolve.umfpack.umfpack' from '...y/linsolve/umfpack/ 
umfpack.pyc'>

And:

****************************************************************
WARNING: cblas module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
   then scipy uses fblas instead of cblas.
****************************************************************


And failures:

======================================================================
FAIL: check loadmat case sparse
----------------------------------------------------------------------
Traceback (most recent call last):
   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages/scipy/io/tests/test_mio.py", line 85, in cc
     self._check_case(name, files, expected)
   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages/scipy/io/tests/test_mio.py", line 80, in  
_check_case
     self._check_level(k_label, expected, matdict[k])
   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages/scipy/io/tests/test_mio.py", line 63, in  
_check_level
     decimal = 5)
   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages/numpy/testing/utils.py", line 232, in  
assert_array_almost_equal
     header='Arrays are not almost equal')
   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages/numpy/testing/utils.py", line 217, in  
assert_array_compare
     assert cond, msg
AssertionError:
Arrays are not almost equal
test sparse; file /Library/Frameworks/Python.framework/Versions/2.5/ 
lib/python2.5/site-packages/scipy/io/tests/./data/ 
testsparse_6.5.1_GLNX86.mat, variable testsparse
(mismatch 46.6666666667%)
  x: array([[  3.03865194e-319,   3.16202013e-322,   1.04346664e-320,
           2.05531309e-320,   2.56123631e-320],
        [  3.16202013e-322,   0.00000000e+000,   0.00000000e+000,...
  y: array([[ 1.,  2.,  3.,  4.,  5.],
        [ 2.,  0.,  0.,  0.,  0.],
        [ 3.,  0.,  0.,  0.,  0.]])

======================================================================
FAIL: check_dot (scipy.lib.tests.test_blas.test_fblas1_simple)
----------------------------------------------------------------------
Traceback (most recent call last):
   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages/scipy/lib/blas/tests/test_blas.py", line 76,  
in check_dot
     assert_almost_equal(f([3j,-4,3-4j],[2,3,1]),-9+2j)
   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/site-packages/numpy/testing/utils.py", line 158, in  
assert_almost_equal
     assert round(abs(desired - actual),decimal) == 0, msg
AssertionError:
Items are not equal:
  ACTUAL: (-1.9988369941711426+4.5782649454569436e-37j)
  DESIRED: (-9+2j)


I am too much of a neophyte to know what to do at this point.

Any help will be appreciated.

D.

On Nov 17, 2007, at 9:41 PM, Robert Kern wrote:

> David Arnold wrote:
>> All,
>>
>> Mac OS X 10.4.11 PPC.
>>
>> Installed python from this:
>>
>> http://www.python.org/ftp/python/2.5.1/python-2.5.1-macosx.dmg
>>
>> Installation finishes with "There were errors installing the
>> software. Please try installing again." Installing again makes no
>> difference, but Python appears to be running:
>>
>> darnold $ python
>> Python 2.5.1 (r251:54869, Apr 18 2007, 22:08:04)
>> [GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin
>> Type "help", "copyright", "credits" or "license" for more  
>> information.
>>>>>
>>
>> Downloaded numpy from:  http://downloads.sourceforge.net/numpy/
>> numpy-1.0.4.tar.gz?use_mirror=internap
>>
>> After unpacking and changing directories, simply did this:
>>
>> numpy-1.0.4 $ python setup.py install
>>
>> Seem to finish OK. But testing according to README.txt resulted in:
>>
>> numpy-1.0.4 $ python -c 'import numpy; numpy.test()'
>> Running from numpy source directory.
>> Traceback (most recent call last):
>>    File "<string>", line 1, in <module>
>> AttributeError: 'module' object has no attribute 'test'
>>
>> Any ideas or advice?
>
> Change to another directory. The numpy/ directory in the source  
> tree is getting
> picked up instead of the installed one. Same thing with scipy.
>
> -- 
> Robert Kern
>
> "I have come to believe that the whole world is an enigma, a  
> harmless enigma
>  that is made terrible by our own mad attempt to interpret it as  
> though it had
>  an underlying truth."
>   -- Umberto Eco
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From youngsu999 at gmail.com  Sun Nov 18 05:50:21 2007
From: youngsu999 at gmail.com (youngsu park)
Date: Sun, 18 Nov 2007 19:50:21 +0900
Subject: [SciPy-user]  scipy 0.6.0 ImportError: cannot import name cscmux
Message-ID: <338357900711180250u1fd08b86ide9a504d664e0cc1@mail.gmail.com>

Hi. Im Youngsu-park ,in POSTECH in south korea.

I had same problem on scipy importing like:

    from scipy import *

.

So I looked at the scipy/sparse folder,

"C:\Python25\Lib\site-packages\scipy\sparse".

To find the reason of error, I open the

"C:\Python25\Lib\site-packages\scipy\sparse\sparse.py", file.

>From line 21 to 26, there was import statement for "sparsetools"



from scipy.sparse.sparsetools import cscmux, csrmux, \

     cootocsr, csrtocoo, cootocsc, csctocoo, csctocsr, csrtocsc, \

     densetocsr, csrtodense, \

     csrmucsr, cscmucsc, \

     csr_plus_csr, csc_plus_csc, csr_minus_csr, csc_minus_csc, \

     csr_elmul_csr, csc_elmul_csc, csr_eldiv_csr, csc_eldiv_csc



I change the working directory to
"C:\Python25\Lib\site-packages\scipy\sparse",

import sparestools and use dir function to see the members of sparsetools



In [16]: import sparsetoo



In [17]: dir(sparsetools)

Out[17]:

['__doc__',  '__file__',  '__name__', '__version__', 'ccootocsc', 'ccscadd',
'ccscextract',

'ccscgetel', 'ccscmucsc', 'ccscmucsr', 'ccscmul', 'ccscmux', 'ccscsetel',
'ccsctocoo',

 'ccsctofull', 'ccsrmucsc', 'ccsrmux', 'cdiatocsc', 'cfulltocsc', 'ctransp',
'dcootocsc',

 'dcscadd', 'dcscextract', 'dcscgetel', 'dcscmucsc', 'dcscmucsr', 'dcscmul',
'dcscmux',

 'dcscsetel', 'dcsctocoo', 'dcsctofull', 'dcsrmucsc', 'dcsrmux',
'ddiatocsc', 'dfulltocsc',

 'dtransp', 'scootocsc', 'scscadd', 'scscextract', 'scscgetel', 'scscmucsc',
'scscmucsr',

 'scscmul', 'scscmux', 'scscsetel', 'scsctocoo', 'scsctofull', 'scsrmucsc',
'scsrmux', 'sdiatocsc',

 'sfulltocsc', 'stransp', 'zcootocsc', 'zcscadd', 'zcscextract',
'zcscgetel', 'zcscmucsc', 'zcscmucsr',

 'zcscmul', 'zcscmux', 'zcscsetel', 'zcsctocoo', 'zcsctofull', 'zcsrmucsc',
'zcsrmux', 'zdiatocsc',

 'zfulltocsc', 'ztransp']



But it is not the members of sparsetools.py

I think it is the members of sparsetools.pyd.



Move sparsetools.pyd to some backup directory.

Then it will works well.



In [21]: from scipy import *



In [22]:
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071118/124293d6/attachment.html>

From youngsu999 at gmail.com  Sun Nov 18 06:06:35 2007
From: youngsu999 at gmail.com (youngsu park)
Date: Sun, 18 Nov 2007 20:06:35 +0900
Subject: [SciPy-user] scipy 0.6.0 ImportError: cannot import name cscmux
In-Reply-To: <338357900711180250u1fd08b86ide9a504d664e0cc1@mail.gmail.com>
References: <338357900711180250u1fd08b86ide9a504d664e0cc1@mail.gmail.com>
Message-ID: <338357900711180306i5eb9e54eif0a34032b93c53ef@mail.gmail.com>

Hi. Im Youngsu-park ,in POSTECH in south korea.

I had same problem on scipy importing like:

    from scipy import *

.

So I looked at the scipy/sparse folder,

"C:\Python25\Lib\site-packages\scipy\sparse ".

To find the reason of error, I open the

"C:\Python25\Lib\site-packages\scipy\sparse\sparse.py", file.

>From line 21 to 26, there was import statement for "sparsetools"



from scipy.sparse.sparsetools import cscmux, csrmux, \

     cootocsr, csrtocoo, cootocsc, csctocoo, csctocsr, csrtocsc, \

     densetocsr, csrtodense, \

     csrmucsr, cscmucsc, \

     csr_plus_csr, csc_plus_csc, csr_minus_csr, csc_minus_csc, \

     csr_elmul_csr, csc_elmul_csc, csr_eldiv_csr, csc_eldiv_csc



I change the working directory to "C:\Python25\Lib\site-packages\scipy\sparse
",

import sparestools and use dir function to see the members of sparsetools



In [16]: import sparsetoo



In [17]: dir(sparsetools)

Out[17]:

['__doc__',  '__file__',  '__name__', '__version__', 'ccootocsc', 'ccscadd',
'ccscextract',

'ccscgetel', 'ccscmucsc', 'ccscmucsr', 'ccscmul', 'ccscmux', 'ccscsetel',
'ccsctocoo',

 'ccsctofull', 'ccsrmucsc', 'ccsrmux', 'cdiatocsc', 'cfulltocsc', 'ctransp',
'dcootocsc',

 'dcscadd', 'dcscextract', 'dcscgetel', 'dcscmucsc', 'dcscmucsr', 'dcscmul',
'dcscmux',

 'dcscsetel', 'dcsctocoo', 'dcsctofull', 'dcsrmucsc', 'dcsrmux',
'ddiatocsc', 'dfulltocsc',

 'dtransp', 'scootocsc', 'scscadd', 'scscextract', 'scscgetel', 'scscmucsc',
'scscmucsr',

 'scscmul', 'scscmux', 'scscsetel', 'scsctocoo', 'scsctofull', 'scsrmucsc',
'scsrmux', 'sdiatocsc',

 'sfulltocsc', 'stransp', 'zcootocsc', 'zcscadd', 'zcscextract',
'zcscgetel', 'zcscmucsc', 'zcscmucsr',

 'zcscmul', 'zcscmux', 'zcscsetel', 'zcsctocoo', 'zcsctofull', 'zcsrmucsc',
'zcsrmux', 'zdiatocsc',

 'zfulltocsc', 'ztransp']



But it is not the members of sparsetools.py

I think it is the members of sparsetools.pyd.



Move sparsetools.pyd to some backup directory.

Then it will works well.



In [21]: from scipy import *



In [22]:



<scipy-user at scipy>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071118/dabdfa25/attachment.html>

From ramercer at gmail.com  Sun Nov 18 12:08:06 2007
From: ramercer at gmail.com (Adam Mercer)
Date: Sun, 18 Nov 2007 12:08:06 -0500
Subject: [SciPy-user] Specifying fortran compiler
In-Reply-To: <54D6FEEC-043D-4B1B-A0E4-B7AB7DC98CA4@physics.mcmaster.ca>
References: <799406d60710281113m7865640ds11d9e66665f09c2b@mail.gmail.com>
	<799406d60710281918j616481a1te3fdcf88c45481a@mail.gmail.com>
	<47254429.4030508@gmail.com>
	<799406d60710282030t77121657s894b6859113700a@mail.gmail.com>
	<54D6FEEC-043D-4B1B-A0E4-B7AB7DC98CA4@physics.mcmaster.ca>
Message-ID: <799406d60711180908j4fbeb53ay4cc5723082b70e9@mail.gmail.com>

On 31/10/2007, David M. Cooke <cookedm at physics.mcmaster.ca> wrote:

> If it's fixed (as I think it is), best to wait for 1.0.4. I suppose
> you could make a symlink to gfortran-mp-42 in the work/ directory, and
> add it to the PATH that'd be used in the Portfile.

This problem goes away with numpy-1.0.4

Cheers

Adam


From berthe.loic at gmail.com  Sun Nov 18 14:08:52 2007
From: berthe.loic at gmail.com (LB)
Date: Sun, 18 Nov 2007 11:08:52 -0800 (PST)
Subject: [SciPy-user] Pb with scipy.stats
Message-ID: <f66d5cdf-5d01-4c2a-9495-ac8c284e6e42@o6g2000hsd.googlegroups.com>

Hi,
I've got some problem with some of the continuous distributions
defined in scipy.stats :

>>> from scipy import stats, __version__
>>> __version__
'0.7.0.dev3440'
>>> beta = stats.betaprime(5., 5.)
>>> beta.pdf( 0.5 )
array(0.68282274043590918)
>>> beta.cdf( 0.5>>> from scipy import stats, __version__
>>> __version__
'0.7.0.dev3440'
>>> beta = stats.betaprime(5., 5.)
>>> beta.pdf( 0.5 )
array(0.68282274043590918)
>>> beta.cdf( 0.5 )
array(0.14484580602550423)

So, PDF and CDF methods work, but not PPF :

>>> beta.ppf( 0.5 )
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/loic/tmp/pycvs/lib/python2.6/site-packages/scipy/stats/
distributions.py", line 112, in ppf
    return self.dist.ppf(q,*self.args,**self.kwds)
  File "/home/loic/tmp/pycvs/lib/python2.6/site-packages/scipy/stats/
distributions.py", line 563, in ppf
    place(output,cond,self._ppf(*goodargs)*scale + loc)
  File "/home/loic/tmp/pycvs/lib/python2.6/site-packages/scipy/stats/
distributions.py", line 382, in _ppf
    return self.vecfunc(q,*args)
  File "/home/loic/tmp/pycvs/lib/python2.6/site-packages/numpy/lib/
function_base.py", line 949, in __call__
    raise ValueError, "mismatch between python function inputs"\
ValueError: mismatch between python function inputs and received
arguments


It seems the same problem appears with at least foldcauchy, foldnorm
and  genexpon

Is there any way to avoid this ?

--
LB


From lev at vpac.org  Sun Nov 18 17:36:25 2007
From: lev at vpac.org (Lev Lafayette)
Date: Mon, 19 Nov 2007 09:36:25 +1100
Subject: [SciPy-user] Installing numpy
In-Reply-To: <7F5B3638-3AF9-4181-9A52-7908DAB7A425@suddenlink.net>
References: <E1EE2D9C-7CA0-44A2-9094-DF743AE8EF2F@suddenlink.net>
	<473FD0A7.3030404@gmail.com>
	<7F5B3638-3AF9-4181-9A52-7908DAB7A425@suddenlink.net>
Message-ID: <1195425385.26337.2.camel@sys09.in.vpac.org>

On Sat, 2007-11-17 at 22:02 -0800, David Arnold wrote:
> Wow. Good catch. Switching directories:
> 
> Ran 692 tests in 1.211s
> 
>  >>>import numpy
>  >>>numpy.test()
> 
> ...
> 
> OK
> <unittest._TextTestResult run=692 errors=0 failures=0>
>  >>>
> 
> 
> Not so lucky with scipy:
> 
> darnold $ python
> Python 2.5.1 (r251:54869, Apr 18 2007, 22:08:04)
> [GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin
> Type "help", "copyright", "credits" or "license" for more information.
>  >>> import scipy
> s>>> scipy.test(level=1)
> 

Is this a 64 bit machine? If so you'll have the same problem with SciPy
as I did... until a very nice member of this list pointed me to the
magic solution.

http://scipy.org/scipy/scipy/changeset/3498

HTH HAND,



-- 
Lev Lafayette <lev at vpac.org>



From david at ar.media.kyoto-u.ac.jp  Sun Nov 18 22:12:05 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 19 Nov 2007 12:12:05 +0900
Subject: [SciPy-user] Adding pysamplerate to scikits ?
Message-ID: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>

Hi,

    Just wanted to know if it would be ok to add pysamplerate to scikits 
? pysamplerate is a (really rough for now) wrapper around SRC for high 
quality audio resampling (using sinc interpolation). It would be LGPL, 
as SRC itself:

http://www.ar.media.kyoto-u.ac.jp/members/david/softwares/pysamplerate/index.html

cheers,

David


From millman at berkeley.edu  Sun Nov 18 23:04:33 2007
From: millman at berkeley.edu (Jarrod Millman)
Date: Sun, 18 Nov 2007 20:04:33 -0800
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
Message-ID: <c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>

On Nov 18, 2007 7:12 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
>     Just wanted to know if it would be ok to add pysamplerate to scikits
> ? pysamplerate is a (really rough for now) wrapper around SRC for high
> quality audio resampling (using sinc interpolation). It would be LGPL,
> as SRC itself:

+1

I would be happy to see you adding a audio resampling scikit.  As you
I am sure you will guess, my only concern is that you rename the
package to at least remove the 'py'.  I would also prefer to see it
named resample over samplerate.

Thanks,

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From millman at berkeley.edu  Sun Nov 18 23:04:33 2007
From: millman at berkeley.edu (Jarrod Millman)
Date: Sun, 18 Nov 2007 20:04:33 -0800
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
Message-ID: <c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>

On Nov 18, 2007 7:12 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
>     Just wanted to know if it would be ok to add pysamplerate to scikits
> ? pysamplerate is a (really rough for now) wrapper around SRC for high
> quality audio resampling (using sinc interpolation). It would be LGPL,
> as SRC itself:

+1

I would be happy to see you adding a audio resampling scikit.  As you
I am sure you will guess, my only concern is that you rename the
package to at least remove the 'py'.  I would also prefer to see it
named resample over samplerate.

Thanks,

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From david at ar.media.kyoto-u.ac.jp  Sun Nov 18 23:06:03 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 19 Nov 2007 13:06:03 +0900
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
Message-ID: <47410BAB.9010902@ar.media.kyoto-u.ac.jp>

Jarrod Millman wrote:
> On Nov 18, 2007 7:12 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
>   
>>     Just wanted to know if it would be ok to add pysamplerate to scikits
>> ? pysamplerate is a (really rough for now) wrapper around SRC for high
>> quality audio resampling (using sinc interpolation). It would be LGPL,
>> as SRC itself:
>>     
>
> +1
>
> I would be happy to see you adding a audio resampling scikit.  As you
> I am sure you will guess, my only concern is that you rename the
> package to at least remove the 'py'.  I would also prefer to see it
> named resample over samplerate.
>   
I indeed thought it would be a 'complain'. No problem, I don't care 
about the name (and it was a bad name anyway).

cheers,

David


From youngsu999 at gmail.com  Mon Nov 19 00:10:34 2007
From: youngsu999 at gmail.com (Youngsu Park)
Date: Mon, 19 Nov 2007 14:10:34 +0900
Subject: [SciPy-user] Pb with scipy.stats
Message-ID: <338357900711182110i6a06756fx1822e0c7dcb606d6@mail.gmail.com>

Hi,
In the scipy/stats directory, there is distribution.py file.
Several distribution is defined in it.
The betaprime is a instance of betaprime_gen that inherit rv_continuous.
The ppf method is not defined In betaprime_gen class.

## Beta Prime
class betaprime_gen(rv_continuous):
    def _rvs(self, a, b):
        u1 = gamma.rvs(a,size=self._size)
        u2 = gamma.rvs(b,size=self._size)
        return (u1 / u2)
    def _pdf(self, x, a, b):
        return 1.0/special.beta(a,b)*x**(a-1.0)/(1+x)**(a+b)
    def _cdf(self, x, a, b):
        x = where(x==1.0, 1.0-1e-6,x)
        return pow(x,a)*special.hyp2f1(a+b,a,1+a,-x)/a/special.beta(a,b)
    def _munp(self, n, a, b):
        if (n == 1.0):
            return where(b > 1, a/(b-1.0), inf)
        elif (n == 2.0):
            return where(b > 2, a*(a+1.0)/((b-2.0)*(b-1.0)), inf)
        elif (n == 3.0):
            return where(b > 3, a*(a+1.0)*(a+2.0)/((b-3.0)*(b-2.0)*(b-1.0)),
                         inf)
        elif (n == 4.0):
            return where(b > 4,
                         a*(a+1.0)*(a+2.0)*(a+3.0)/((b-4.0)*(b-3.0) \
                                                    *(b-2.0)*(b-1.0)), inf)
        else:
            raise NotImplementedError
betaprime = betaprime_gen(a=0.0, b=500.0, name='betaprime', shapes='a,b',
                          extradoc="""~~~""")



so it uses the rv_continuous class's method.

    def ppf(self,q,*args,**kwds):
        """Percent point function (inverse of cdf) at q of the given RV.

        *args
        =====
          The shape parameter(s) for the distribution (see docstring of the
           instance object for more information)

        **kwds
        ======
        loc   - location parameter (default=0)
        scale - scale parameter (default=1)
        """
        loc,scale=map(kwds.get,['loc','scale'])
        args, loc, scale = self.__fix_loc_scale(args, loc, scale)
        q,loc,scale = map(arr,(q,loc,scale))
        args = tuple(map(arr,args))
        cond0 = self._argcheck(*args) & (scale > 0) & (loc==loc)
        cond1 = (q > 0) & (q < 1)
        cond2 = (q==1) & cond0
        cond = cond0 & cond1
        output = valarray(shape(cond),value=self.a*scale + loc)
        place(output,(1-cond0)+(1-cond1)*(q!=0.0), self.badvalue)
        place(output,cond2,self.b*scale + loc)
        goodargs = argsreduce(cond, *((q,)+args+(scale,loc)))
        scale, loc, goodargs = goodargs[-2], goodargs[-1], goodargs[:-2]
        place(output,cond,self._ppf(*goodargs)*scale + loc)
         ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
        return output
....
    sgf = vectorize
    self.vecfunc = sgf(self._ppf_single_call,otypes='d')
...
    def _ppf(self, q, *args):
        return self.vecfunc(q,*args)
        ~~~~~~~~~~~~~~~~~~~~
...    def _ppf_single_call(self, q, *args):
         return scipy.optimize.brentq(self._ppf_to_solve, self.xa,
self.xb, args=(q,)+args, xtol=self.xtol)
I'm not sure but I think that the problem comes from the vectorization
of defalut _ppf_single_call funstion.
But I don'k know how to fix it.
Is there someone to fix it?


From pav at iki.fi  Mon Nov 19 03:14:02 2007
From: pav at iki.fi (Pauli Virtanen)
Date: Mon, 19 Nov 2007 08:14:02 +0000 (UTC)
Subject: [SciPy-user] ode on complex systems
References: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
	<1195226348.24691.2.camel@localhost> <web-92145595@uni-stuttgart.de>
	<473DD6AC.3070207@noaa.gov>
	<80c99e790711161108k6fbdd4b4sd1db90932d1771d6@mail.gmail.com>
	<fhkv9a$q1g$1@ger.gmane.org>
Message-ID: <fhrgka$lc9$1@ger.gmane.org>

Fri, 16 Nov 2007 20:41:14 +0000, Pauli Virtanen wrote:

> Fri, 16 Nov 2007 20:08:22 +0100, lorenzo bolla wrote:
>> I quickly tried to compile scipy with the patch, but it failed (on
>> cygwin). I had no time to investigate the problem further, but I'll let
>> you know as soon as possible more details. L.
> 
> This may be a cygwin-related issue, 

Ok, it seems that someone else is also having problems on Cygwin (see the 
post titled "Problems compiling Scipy with Cygwin"): the VODE module even 
without patches fails to compile. Did you check whether you are compile 
an unpatched Scipy?

-- 
Pauli Virtanen



From lev at columbia.edu  Mon Nov 19 09:20:43 2007
From: lev at columbia.edu (Lev Givon)
Date: Mon, 19 Nov 2007 09:20:43 -0500
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
Message-ID: <20071119142042.GA21903@localhost.cc.columbia.edu>

Received from Jarrod Millman on Sun, Nov 18, 2007 at 11:04:33PM EST:
> On Nov 18, 2007 7:12 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
> >     Just wanted to know if it would be ok to add pysamplerate to scikits
> > ? pysamplerate is a (really rough for now) wrapper around SRC for high
> > quality audio resampling (using sinc interpolation). It would be LGPL,
> > as SRC itself:

SRC is licensed under the GPL, not the LGPL:

http://www.mega-nerd.com/SRC/license.html

Offhand, I am not sure how this impacts the distribution of David's
wrapper under the BSD license.

> 
> +1
> 
> I would be happy to see you adding a audio resampling scikit.  As you

Me too; I was actually looking for something like this recently.

> I am sure you will guess, my only concern is that you rename the
> package to at least remove the 'py'.  I would also prefer to see it
> named resample over samplerate.
> 
> Thanks,
> 
> -- 
> Jarrod Millman

Although I agree that there are better names than "samplerate", it
might be preferable to retain the name of the underlying library (or
perhaps choose something else entirely) because there is another free
resampling library called (lib)resample:

http://www-ccrma.stanford.edu/~jos/resample/

Given my previous observation regarding SRC's license, it is worth
noting that this library is licensed under the LGPL (supposedly
because the author found SRC's license problematic for his purposes).

							L.G.



From lbolla at gmail.com  Mon Nov 19 10:35:36 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Mon, 19 Nov 2007 16:35:36 +0100
Subject: [SciPy-user] ode on complex systems
In-Reply-To: <fhrgka$lc9$1@ger.gmane.org>
References: <80c99e790711160653v58f2b65o68a52c422e4623fc@mail.gmail.com>
	<1195226348.24691.2.camel@localhost> <web-92145595@uni-stuttgart.de>
	<473DD6AC.3070207@noaa.gov>
	<80c99e790711161108k6fbdd4b4sd1db90932d1771d6@mail.gmail.com>
	<fhkv9a$q1g$1@ger.gmane.org> <fhrgka$lc9$1@ger.gmane.org>
Message-ID: <80c99e790711190735w1a9f2719o2b1843506f05729b@mail.gmail.com>

I managed to compile scipy with the zvode.patch.

Just one observation: doing
$> patch < zvode.patch
in the scipy/scipy/integrate directory is not enough: you also need to move
zvode.f in odepack/ directory and the others *.f in linpack_lite/ directory.
That is what I missed last time.

Thank you again,
Lorenzo.


On 11/19/07, Pauli Virtanen <pav at iki.fi> wrote:
>
> Fri, 16 Nov 2007 20:41:14 +0000, Pauli Virtanen wrote:
>
> > Fri, 16 Nov 2007 20:08:22 +0100, lorenzo bolla wrote:
> >> I quickly tried to compile scipy with the patch, but it failed (on
> >> cygwin). I had no time to investigate the problem further, but I'll let
> >> you know as soon as possible more details. L.
> >
> > This may be a cygwin-related issue,
>
> Ok, it seems that someone else is also having problems on Cygwin (see the
> post titled "Problems compiling Scipy with Cygwin"): the VODE module even
> without patches fails to compile. Did you check whether you are compile
> an unpatched Scipy?
>
> --
> Pauli Virtanen
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071119/06fd128d/attachment.html>

From odonnems at yahoo.com  Mon Nov 19 11:06:03 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Mon, 19 Nov 2007 08:06:03 -0800 (PST)
Subject: [SciPy-user] weave and compiler install help
Message-ID: <271728.69580.qm@web58012.mail.re3.yahoo.com>

I am trying to compile some inline c++ code inside python using weave.
I always get a similar problem where the compiled file cannot be found
(see below). I am not sure if the problem is with the compiler
or something else. I am a new user of scipy and a novice with python so
I would appreciate any direction someone can give me because I have not
been able to figure out a work around.

Thank you,
Michael

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
I have the following related applications installed:
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Python:
2.4.1 (I am using an older version because this is what ESRI GIS application 
 supports)
 
Scipy:
scipy-0.6.0.win32-py2.4.exe

MinGW:
MinGW-5.1.3.exe
gcc-4.1.2-mingw-setup.exe

cygwin:
3.2.25

OS:
Windows XP SP2


+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
When I test the installation of Weave I get the following output:
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>> weave.test()
  Found 1 tests for scipy.weave.ast_tools
  Found 2 tests for scipy.weave.blitz_tools
  Found 9 tests for scipy.weave.build_tools
  Found 0 tests for scipy.weave.c_spec
  Found 26 tests for scipy.weave.catalog
building extensions here: c:\docume~1\michael\locals~1\temp\Michael\python24_compiled\m3
  Found 1 tests for scipy.weave.ext_tools
  Found 0 tests for scipy.weave.inline_tools
  Found 74 tests for scipy.weave.size_check
  Found 16 tests for scipy.weave.slice_handler
  Found 3 tests for
 scipy.weave.standard_array_spec
  Found 0 tests for __main__
...warning: specified build_dir '_bad_path_' does not exist or is not writable. Trying default locations
.....warning: specified build_dir '_bad_path_' does not exist or is not writable. Trying default locations
............................removing 'c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test' (and everything under it)
error
removing c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test:
c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test\win3224compiled_catalog:
Permission denied
error removing
c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test:
c:\docume~1\michael\locals~1\temp\tmpdqudhmcat_test: Directory not empty
.removing 'c:\docume~1\michael\locals~1\temp\tmpw144aycat_test' (and everything under
 it)
...............................................................................................
----------------------------------------------------------------------
Ran 132 tests in 2.625s

OK
<unittest.TextTestRunner object at 0x020D6F70>


+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
When I try to test the following script or any other script I get the following message:
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
def prod(m, v):
    #C++ version
    nrows, ncolumns = m.shape

    res = numpy.zeros((nrows, ncolumns), float)
    code = r"""
    for (int i=0; i<nrows; i++)
    {
        for (int j=0; j<ncolumns; j++)
       
 {
            res(i) += m(i,j)*v(j);
        }
    }
    """

   
#err = weave.inline(code,['nrows', 'ncolumns', 'res', 'm', 'v'],
type_converters=converters.blitz, compiler='mingw32', verbose=2)  
    err = weave.inline(code,['nrows', 'ncolumns', 'res', 'm', 'v'], verbose=2)

<weave: compiling>
running build_ext
running build_src
building extension "sc_045cfaf40ef1a0738b1066aaa886a55d7" sources
customize Mingw32CCompiler
customize Mingw32CCompiler using build_ext
customize Mingw32CCompiler
customize Mingw32CCompiler using build_ext
building 'sc_045cfaf40ef1a0738b1066aaa886a55d7' extension
compiling C++ sources
C compiler: g++ -mno-cygwin -O2 -Wall

compile
options: '-IC:\Python24\lib\site-packages\scipy\weave
-IC:\Python24\lib\site-packages\scipy\weave\scxx
-IC:\Python24\lib\site-packages\numpy\core\include
-IC:\Python24\include -IC:\Python24\PC -c'
g++
-mno-cygwin -O2 -Wall -IC:\Python24\lib\site-packages\scipy\weave
-IC:\Python24\lib\site-packages\scipy\weave\scxx
-IC:\Python24\lib\site-packages\numpy\core\include
-IC:\Python24\include -IC:\Python24\PC -c
c:\docume~1\michael\locals~1\temp\Michael\python24_compiled\sc_045cfaf40ef1a0738b1066aaa886a55d7.cpp
-o
c:\docume~1\michael\locals~1\temp\Michael\python24_intermediate\compiler_c8350d870e6c54e8f29dd7094c2bfb45\Release\docume~1\michael\locals~1\temp\michael\python24_compiled\sc_045cfaf40ef1a0738b1066aaa886a55d7.o
Traceback (most recent call last):
  File "C:\Python24\Lib\site-packages\pythonwin\pywin\framework\scriptutils.py", line 310, in RunScript
    exec codeObject in __main__.__dict__
  File "C:\Documents and Settings\Michael\Application
 Data\ESRI\ArcToolbox\scripts\test_weave.py", line 179, in ?
    main()
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 170, in main
    prod(m, v)
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 44, in prod
    err = weave.inline(code,['nrows', 'ncolumns', 'res', 'm', 'v'], verbose=2)
  File "C:\Python24\Lib\site-packages\scipy\weave\inline_tools.py", line 338, in inline
    auto_downcast = auto_downcast,
  File "C:\Python24\Lib\site-packages\scipy\weave\inline_tools.py", line 447, in compile_function
    verbose=verbose, **kw)
  File "C:\Python24\Lib\site-packages\scipy\weave\ext_tools.py", line 365, in compile
    verbose = verbose, **kw)
  File
 "C:\Python24\Lib\site-packages\scipy\weave\build_tools.py", line 269, in build_extension
    setup(name = module_name, ext_modules = [ext],verbose=verb)
  File "C:\Python24\lib\site-packages\numpy\distutils\core.py", line 173, in setup
    return old_setup(**new_attr)
  File "C:\Python24\lib\distutils\core.py", line 159, in setup
    raise SystemExit, error
CompileError: error: c:\docume~1\michael\locals~1\temp\tmp2dwbkp: No such file or directory

This file is created and automatically opens in python:
import os
import sys
sys.path.insert(0,'C:\\Python24\\lib\\site-packages\\numpy\\distutils')
from exec_command import exec_command
del sys.path[0]
cmd = ['g++', '-mno-cygwin', '-O2', '-Wall', '-IC:\\Python24\\lib\\site-packages\\scipy\\weave', '-IC:\\Python24\\lib\\site-packages\\scipy\\weave\\scxx', '-IC:\\Python24\\lib\\site-packages\\scipy\\weave\\blitz', '-IC:\\Python24\\lib\\site-packages\\numpy\\core\\include', '-IC:\\Python24\\include', '-IC:\\Python24\\PC', '-c', 'c:\\docume~1\\michael\\locals~1\\temp\\Michael\\python24_compiled\\sc_e6928d48a5ee12fb55c3b29ed6bf49ef4.cpp', '-o', 'c:\\docume~1\\michael\\locals~1\\temp\\Michael\\python24_intermediate\\compiler_921933e1c4e3c013306c4ed6f4c15144\\Release\\docume~1\\michael\\locals~1\\temp\\michael\\python24_compiled\\sc_e6928d48a5ee12fb55c3b29ed6bf49ef4.o']
os.environ = {'TMP': 'C:\\WINDOWS\\TEMP', 'COMPUTERNAME': 'ALDER', 'USERDOMAIN': 'ALDER', 'ARCHOME': 'C:\\arcgis\\arcexe9x', 'COMMONPROGRAMFILES': 'C:\\Program Files\\Common Files', 'PROCESSOR_IDENTIFIER': 'x86 Family 6 Model 13 Stepping 8, GenuineIntel', 'PROGRAMFILES': 'C:\\Program Files', 'PROCESSOR_REVISION': '0d08', 'PATH': 'c:\\mingw\\bin;C:\\Program Files\\Common Files\\Roxio Shared\\DLLShared;C:\\arcgis\\arcexe9x\\bin;C:\\Program Files\\FWTools1.2.2\\pymod;C:\\Program Files\\R\\R-2.5.1\\bin;C:\\Documents and Settings\\Michael\\My Documents\\mod_web\\ems_web\\web-content\\mapserver-4.8.1bin;C:\\cygwin\\bin;C:\\program files\\imagemagick-6.3.6-q16;C:\\Python24;C:\\Program Files\\gs\\gs8.51\\lib;C:\\Program Files\\gnuplot\\bin;C:\\ModisTools\\MRT\\bin;C:\\ModisTools\\LDOPE\\LDOPE_Win_bin\\ANCILLARY;C:\\Documents and Settings\\Michael\\My Documents\\my_home\\py_scripting\\src\\tools;C:\\Program
 Files\\HEG\\HEG_Win\\bin;C:\\Perl\\site\\bin;C:\\Perl\\bin;C:\\WINDOWS\\system32;C:\\WINDOWS;C:\\WINDOWS\\System32\\Wbem;C:\\Program Files\\Microsoft SQL Server\\80\\Tools\\Binn\\;C:\\Program Files\\MySQL\\MySQL Server 5.0\\bin;C:\\Program Files\\QuickTime\\QTSystem;C:\\WINDOWS\\system32\\config\\systemprofile\\Local Settings\\Temp;C:\\ModisTools\\MRT\\bin; C:\\ModisTools\\LDOPE\\LDOPE_Win_bin\\ANCILLARY;C:\\Program Files\\R\\R-2.5.1\\bin;;c:\\Program Files\\HEG\\HEG_Win\\bin;c:\\Program Files\\HEG\\HEG_Win\\bin;c:\\Program Files\\HEG\\HEG_Win\\bin', 'SYSTEMROOT': 'C:\\WINDOWS', 'ATHOME': 'C:\\arcgis\\arcexe9x\\arctools', 'SCRIPTING': 'C:\\Documents and Settings\\Michael\\My Documents\\my_home\\py_scripting', 'ARCHOME_USER': 'C:\\arcgis\\arcexe9x', 'TEMP': 'C:\\DOCUME~1\\Michael\\LOCALS~1\\Temp', 'PROCESSOR_ARCHITECTURE': 'x86', 'ANCPATH': 'C:\\ModisTools\\LDOPE\\LDOPE_Win_bin\\ANCILLARY', 'ALLUSERSPROFILE': 'C:\\Documents and Settings\\All Users',
 'SESSIONNAME': 'Console', 'ARCINFOFONTSIZE': '8', 'HOMEPATH': '\\Documents and Settings\\Michael', 'USERNAME': 'Michael', 'LOGONSERVER': '\\\\ALDER', 'COMSPEC': 'C:\\WINDOWS\\system32\\cmd.exe', 'QTJAVA': 'C:\\Program Files\\Java\\jre1.6.0_01\\lib\\ext\\QTJava.zip', 'PYTHONPATH': 'C:\\Program Files\\ArcGIS\\bin;C:\\Python24\\Lib\\site-packages\\gdal;C:\\Program Files\\R\\R-2.5.1\\bin;C:\\Documents and Settings\\Michael\\My Documents\\my_home\\py_scripting\\src\\tools;C:\\Documents and Settings\\Michael\\My Documents\\my_home\\py_scripting\\lib;C:\\WINDOWS\\system32\\config\\systemprofile\\Local Settings\\Temp;C:\\cygwin\\bin;C:\\mingw\\bin;c:\\mingw\\bin;C:\\Program Files\\Common Files\\Roxio Shared\\DLLShared;C:\\arcgis\\arcexe9x\\bin;C:\\Program Files\\FWTools1.2.2\\pymod;C:\\Program Files\\R\\R-2.5.1\\bin;C:\\Documents and Settings\\Michael\\My Documents\\mod_web\\ems_web\\web-content\\mapserver-4.8.1bin;C:\\cygwin\\bin;C:\\program
 files\\imagemagick-6.3.6-q16;C:\\Python24;C:\\Program Files\\gs\\gs8.51\\lib;C:\\Program Files\\gnuplot\\bin;C:\\ModisTools\\MRT\\bin;C:\\ModisTools\\LDOPE\\LDOPE_Win_bin\\ANCILLARY;C:\\Documents and Settings\\Michael\\My Documents\\my_home\\py_scripting\\src\\tools;C:\\Program Files\\HEG\\HEG_Win\\bin;C:\\Perl\\site\\bin;C:\\Perl\\bin;C:\\WINDOWS\\system32;C:\\WINDOWS;C:\\WINDOWS\\System32\\Wbem;C:\\Program Files\\Microsoft SQL Server\\80\\Tools\\Binn\\;C:\\Program Files\\MySQL\\MySQL Server 5.0\\bin;C:\\Program Files\\QuickTime\\QTSystem;%USERPROFILE%\\Local Settings\\Temp;', 'CLASSPATH': '.;C:\\Program Files\\Java\\jre1.6.0_01\\lib\\ext\\QTJava.zip', 'PLAT': 'win32', 'PYTHONSRC': 'C:\\Python24', 'PGSHOME': 'c:\\Program Files\\HEG\\HEG_Win\\TOOLKIT_MTD', 'PATHEXT': '.COM;.EXE;.BAT;.CMD;.VBS;.VBE;.JS;.JSE;.WSF;.WSH;.py;.pdf', 'CLIENTNAME': 'Console', 'FP_NO_HOST_CHECK': 'NO', 'WINDIR': 'C:\\WINDOWS', 'APPDATA': 'C:\\Documents and
 Settings\\Michael\\Application Data', 'HOMEDRIVE': 'C:', 'ARCINFOFONTNAME': 'Courier New', 'SYSTEMDRIVE': 'C:', 'NUMBER_OF_PROCESSORS': '1', 'ARCGISHOME': 'C:\\Program Files\\ArcGIS\\', 'MRTBINDIR': 'c:\\Program Files\\HEG\\HEG_Win\\bin', 'PROCESSOR_LEVEL': '6', 'MRTDATADIR': 'C:\\ModisTools\\MRT\\data', 'OS': 'Windows_NT', '__COMPAT_LAYER': 'EnableNXShowUI ', 'USERPROFILE': 'C:\\Documents and Settings\\Michael'}
s,o = exec_command(cmd, _with_python=0, **{})
f=open('c:\\docume~1\\michael\\locals~1\\temp\\tmp0gk0g_',"w")
f.write(str(s))
f.close()
f=open('c:\\docume~1\\michael\\locals~1\\temp\\tmpadrrdg',"w")
f.write(o)
f.close()





      ____________________________________________________________________________________
Get easy, one-click access to your favorites. 
Make Yahoo! your homepage.
http://www.yahoo.com/r/hs 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071119/772c7861/attachment.html>

From keflavich at gmail.com  Mon Nov 19 14:05:53 2007
From: keflavich at gmail.com (Keflavich)
Date: Mon, 19 Nov 2007 11:05:53 -0800 (PST)
Subject: [SciPy-user] Pylab plotting: math range errors unrecoverable
In-Reply-To: <B3E9751A-B518-4ABB-9DB7-D0BFAB7E2290@physics.ubc.ca>
References: <fceb3d4a0711151107t7fe15b1fv4138dc70b61d9b75@mail.gmail.com> 
	<B3E9751A-B518-4ABB-9DB7-D0BFAB7E2290@physics.ubc.ca>
Message-ID: <25d9ffcb-c0d3-4268-bd47-612b5273a095@s6g2000prc.googlegroups.com>

Thanks Michael.  Any thoughts on automatically masking bad values?

Adam

On Nov 15, 12:19 pm, Michael McNeil Forbes <mfor... at physics.ubc.ca>
wrote:
> On 15 Nov 2007, at 11:07 AM, Adam Ginsburg wrote:> 2. when I get the above error, then try to plot something WITHOUT
> > infinities, it fails with the exact same error, and when I ran the
> > debugger it claimed that meanv was still inf.  Is there any way to fix
> > that short of quitting and reentering the python command line?
>
> You can usually just clear the plot with clf().  I think the plot
> keeps the old data and then tries to redisplay it each time, giving
> errors over and over until you clear the figure.
>
> Michael.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-u... at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user


From darkwind.87 at gmail.com  Mon Nov 19 19:31:01 2007
From: darkwind.87 at gmail.com (Dark Wind)
Date: Mon, 19 Nov 2007 16:31:01 -0800
Subject: [SciPy-user] Problem with NLP in OpenOpt
In-Reply-To: <8793ae6e0711150854g7b78db7ex4738dd0dd00cbe13@mail.gmail.com>
References: <8a6035a00711150056r62c30b6cyd5fb4052eaa3f8b8@mail.gmail.com>
	<8793ae6e0711150854g7b78db7ex4738dd0dd00cbe13@mail.gmail.com>
Message-ID: <8a6035a00711191631n7994c989g435b96d3e8aba6ab@mail.gmail.com>

Thank a lot. Scipy_Cobyla worked, but I am not getting how exactly to
install ALGENCAN on windows? Do I need cygwin to install it on
windows?

thanks

On Nov 15, 2007 8:54 AM, Dominique Orban <dominique.orban at gmail.com> wrote:
> On 11/15/07, Dark Wind <darkwind.87 at gmail.com> wrote:
> > The initial value x0 satisfies the constraints hi's but i am still
> > getting that there is no feasible solution.
> >
> > Can anyone help me with this?
>
> Cobyla is an infeasible method, meaning that in order to identifiy a
> solution to your problem, it will venture outside of the feasible set.
> In your example, it wasn't able to reach a feasible solution within
> the given limits (max number of iterations, tolerances, ...) It is not
> because their is a feasible point that the code will be able to find a
> solution; in this case, it got stuck before reaching the feasible set
> again.
>
> You can either re-solve your problem with looser tolerances, which is
> not very satisfactory, you can re-solve from a different starting
> point (which doesn't have to be feasible), or you can try another
> solver (as Dmitrey showed in his response).
>
> Dominique
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From david at ar.media.kyoto-u.ac.jp  Mon Nov 19 22:28:42 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 20 Nov 2007 12:28:42 +0900
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <20071119142042.GA21903@localhost.cc.columbia.edu>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
	<20071119142042.GA21903@localhost.cc.columbia.edu>
Message-ID: <4742546A.7090200@ar.media.kyoto-u.ac.jp>

Lev Givon wrote:
> Received from Jarrod Millman on Sun, Nov 18, 2007 at 11:04:33PM EST:
>> On Nov 18, 2007 7:12 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
>>>     Just wanted to know if it would be ok to add pysamplerate to scikits
>>> ? pysamplerate is a (really rough for now) wrapper around SRC for high
>>> quality audio resampling (using sinc interpolation). It would be LGPL,
>>> as SRC itself:
>
> SRC is licensed under the GPL, not the LGPL:
>
> http://www.mega-nerd.com/SRC/license.html
>
> Offhand, I am not sure how this impacts the distribution of David's
> wrapper under the BSD license.
Indeed, I was mixing SRC and sndfile licenses (both are by Erik Castro 
de Lopo). The plan is to do as Robert Kern suggested: release the 
wrapper with the same license as the wrapped library (GPL here).
>
>> +1
>>
>> I would be happy to see you adding a audio resampling scikit.  As you
>
> Me too; I was actually looking for something like this recently.
If I had more time, I would implement one in pure python, using 
filterbank (which is how matlab does it, AFAIK, and is worthwile as a 
'pedagogical' tool). But well, I don't have this time :)

cheers,

David


From xavier.barthelemy at cmla.ens-cachan.fr  Tue Nov 20 06:21:29 2007
From: xavier.barthelemy at cmla.ens-cachan.fr (Xavier Barthelemy)
Date: Tue, 20 Nov 2007 12:21:29 +0100
Subject: [SciPy-user] organizing data
In-Reply-To: <4742546A.7090200@ar.media.kyoto-u.ac.jp>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>	<20071119142042.GA21903@localhost.cc.columbia.edu>
	<4742546A.7090200@ar.media.kyoto-u.ac.jp>
Message-ID: <4742C339.8010308@cmla.ens-cachan.fr>

Hi everyone!

I'm quite new to python, and I am breaking my mind on this:
I would like to organize data so I can plot them.

My data are the variation of 11 parameters, with 18 values for each. so 
you can see them as an hypercube of 11 dimension. My goal is the 
possibility to plot with gnuplot in 2D (or 3D). So I have to class my 
data by slices of 9 (or 8) parameters constant.
That's my problem: it is not really an "hyper"matrix of data because i 
did not discretized the space parameters equally for each parameters.
So i think i have to class and then recursively cut into the mass so I 
can have small arrays I can plot with the gnuplot.py

As I said, i am quite new in pythonizing, and I sure can use a little 
help. I have spent much time on this, and I can't figure it out how.


so:

from numpy import *
from scipy import *
from operator import itemgetter
import Gnuplot, Gnuplot.PlotItems, Gnuplot.funcutils


#I read the data

def LireFich(Data):
     donnees= io.read_array(Data,columns=(0,-1),lines=(0,-1),atype='d')
     return donnees


def Couper(liste,col):

#Then I class the whole array (sorted can class by each and every 
columns just add a coma and a new col)

     longueur = len(liste)
     listeSansDoublon=[]
     liste2=array(sorted(liste,key=itemgetter(col)))

#Then i create a list of each values of the parameters so i can cut by 
each constant "pieces"
     for i in liste2[:,col] :
	 if listeSansDoublon.count(i) == 0:
		 listeSansDoublon.append(i)
     longueur2=len(listeSansDoublon)

and then... I don't know


I have guessed the array_split function

	OuCouper=[]
	k=0
	for i in listeSansDoublon:
		j = array(where(i == liste2[:,col]))
		jj= where(i == liste2[:,col])
		A=min(j)
		B=max(j)
		OuCouper[k,1]=A
		OuCouper[k,2]=B
	
#	essai=array_split(liste2,min(j)--max(j))



but actually it doesn't work, as i have read the F*M*.


i have also think that I can only manipulate masks ans apply the final 
recursively calculated mask on the data and then reduce to have the 
correct form for gnuplot.

Thanks for your help if you can help me. I really passed much time on 
this, and i still can't figure it out how.


Xavier



From ramercer at gmail.com  Tue Nov 20 09:03:35 2007
From: ramercer at gmail.com (Adam Mercer)
Date: Tue, 20 Nov 2007 09:03:35 -0500
Subject: [SciPy-user] Building with a non-standard gfortran
Message-ID: <799406d60711200603t8e451cah2e13add2bd794a11@mail.gmail.com>

Hi

Using MacPorts, the gfortran compiler, from gcc-4.2.x, is installed as
gfortran-mp-4.2.  I have been trying to build scipy with this compiler
and found that that numpy I need to build and install with

$ python setup.py config_fc --fcompiler gnu95 \
    --f77exec /opt/local/bin/gfortran-mp-4.2 \
    --f90exec /opt/local/bin/gfortran-mp-4.2 build
$ python setup.py install --prefix=${NUMPY_LOCATION}

however using the above for scipy fails on the install phase with it
being unable to locate a fortran compiler.  I therefore need to
install scipy with

$ python setup.py config_fc --fcompiler gnu95 \
    --f77exec /opt/local/bin/gfortran-mp-4.2 \
    --f90exec /opt/local/bin/gfortran-mp-4.2 install --prefix=${SCIPY_LOCATION}

Is this expected, as I would have expected install to inherit the
options that where used during the build phase?

Cheers

Adam

PS: I'm using numpy-1.0.4 and scipy-0.6.0


From aisaac at american.edu  Tue Nov 20 09:51:59 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Tue, 20 Nov 2007 09:51:59 -0500
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <4742546A.7090200@ar.media.kyoto-u.ac.jp>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com><20071119142042.GA21903@localhost.cc.columbia.edu><4742546A.7090200@ar.media.kyoto-u.ac.jp>
Message-ID: <Mahogany-0.67.0-1296-20071120-095159.00@american.edu>

On Tue, 20 Nov 2007, David Cournapeau apparently wrote:
> I was mixing SRC and sndfile licenses (both are by Erik 
> Castro de Lopo). The plan is to do as Robert Kern 
> suggested: release the wrapper with the same license as 
> the wrapped library (GPL here). 

If you anticipate a reduction in usefulness from this,
why not explain that and ask Erik for a release under the LGPL?
Licenses are not a "given", especially when the number of 
developers is small enough to feasibly agree on an 
additional license.

Cheers,
Alan Isaac





From timmichelsen at gmx-topmail.de  Tue Nov 20 11:47:21 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Tue, 20 Nov 2007 16:47:21 +0000 (UTC)
Subject: [SciPy-user] help on array manipulation
Message-ID: <loom.20071120T163815-497@post.gmane.org>

Hello,

I have the following array:

In [240]: z
Out[240]:
array([[ 15.,   3.,   7.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
          0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
          0.,   0.,   0.],
       [ 16.,   3.,   7.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
          0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
          0.,   0.,   0.],
       [ 17.,   3.,   7.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
          0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
          0.,   0.,   0.]])

How do I get it be become like:

15 3
15 7
15 0
...
16 3
16 7
...
17 3


Is there a special function I can use for this or do I need to build various for
loops to achive this?

I mean this example has few data. But what if the amount increases?

I am stuck here.

I'd be very glad if anyone could point be to a good tutorial for array
manipulation. Scipy Cookbook is not enough.
Well, I also found :
http://numpy.scipy.org/numpydoc
and
http://www.penzilla.net/tutorials/python/numeric

Thanks in advance,
Timmie


###
# Code for the array:

###
z = numpy.zeros((3,25))
z = numpy.zeros((3,25))
for i in xrange(z.shape[1]):
    z[:,0] = 1
    z[:,1] = 3
    z[:,2] = 7
    

for i in xrange(z.shape[0]):
    z[0,0] = 15
    z[1,0] = 16
    z[2,0] = 17



From timmichelsen at gmx-topmail.de  Tue Nov 20 17:58:07 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Tue, 20 Nov 2007 22:58:07 +0000 (UTC)
Subject: [SciPy-user] reading data with gaps and converting
Message-ID: <loom.20071120T225033-310@post.gmane.org>

Hello,
I encountered another difficulty converting my files from one
format to the other.

Below you may find a simplified example.

When I try to load for instance row 1 on the input file the pylab.load
or scipy.io.read_array complains that the index is out of range or
that a sequence cannot be converted into a array.
I assume that the problem are the cells without any data.
The file I use was exported from Excel.
There may be some of you who have had to deal with similar format
conversion issues. Therefore I would be very happy someone could indicate
me a method how to read such structured data efficiently into a array.

Thanks in advance.

Kind regards,
Tim 


reduced example data:

input file structure

     parameter1    parameter1    parameter2    parameter2    
year       month      day    hour1     hour2     hour1     hour2     
2007     11     20       8         2
2007     11     21         9         3         4         4
2007     11     22         2         8         4         8
2007     11     23         3         3         6        51
2007     11     24       1        72        34
2007     11     25         7         9        34        24
2007     11     26         6        65        12         1
2007     11     27         7         0        19         2
2007     11     28      44        21         2
       
                                                                           
                                                                           
output file structure                   
                                                      
                                                                           
       
year       month      day    hour      parameter1    parameter2    
       2007     11     20         1         nodata    2
       2007     11     20         2         8        nodata    
       2007     11     21         1         9         4
       2007     11     21         2         3         4
       2007     11     22         1         2         4
       2007     11     22         2         8         8



From stuwilkins at mac.com  Tue Nov 20 18:09:37 2007
From: stuwilkins at mac.com (Stuart Wilkins)
Date: Tue, 20 Nov 2007 15:09:37 -0800
Subject: [SciPy-user] Questions on leastsq
Message-ID: <7D7AE45E-0116-1000-8386-17482B0ECB9E-Webmail-10020@mac.com>

Hi, 

I am having problems with the least squares fitting. This may or may not be a bug.

I am trying to get an idea of the errors from fitting data. I am usualy fitting to Gaussian Lorentzian, Lor**2 etc. When I try to compute the errors from

sigma = sqrt(diag(cov_x))

where cov_x is the covariance matrix I get results which are wildly different than anything before. Including some code I had a long time ago in Matlab.

Does anyone have any idea if the form of cov_x is correct or any success in using this?

Thanks,
Stuart


From anand.prabhakar.patil at gmail.com  Tue Nov 20 18:30:01 2007
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Tue, 20 Nov 2007 15:30:01 -0800
Subject: [SciPy-user] Sparse matrix advice
Message-ID: <2bc7a5a50711201530l25b3b813j673be7620c1e8154@mail.gmail.com>

Hi all,

I'm looking for a 'row-ish' sparse matrix format that I can operate on
(including making new nonzero elements) quickly in C++ or Pyrex. Could
anyone recommend one?

The LIL format looks nice because it can be expanded without any
copying, but I'm worried that having to call PyList_GetItem (or
whatever it is), inc/decreffing and converting from Python to C
integers will take up a lot of time. PyTrilinos' Epetra looks pretty
cool but I'm not sure how much distributed memory stuff I'll be doing
and it would take me a long time to understand.

I'm thinking that a good solution might be Pyrex wrappers for matrices
represented as C++ vectors of rows, which are themselves vectors;
would this actually be a good solution, and is anything like this out
there already?

I'm a total sparse linear algebra neophyte, so if I'm missing some
boat entirely please help me get a clue.

Many thanks in advance,
Anand Patil


From david at ar.media.kyoto-u.ac.jp  Tue Nov 20 21:26:38 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 21 Nov 2007 11:26:38 +0900
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <Mahogany-0.67.0-1296-20071120-095159.00@american.edu>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com><20071119142042.GA21903@localhost.cc.columbia.edu><4742546A.7090200@ar.media.kyoto-u.ac.jp>
	<Mahogany-0.67.0-1296-20071120-095159.00@american.edu>
Message-ID: <4743975E.10705@ar.media.kyoto-u.ac.jp>

Alan G Isaac wrote:
> On Tue, 20 Nov 2007, David Cournapeau apparently wrote:
>> I was mixing SRC and sndfile licenses (both are by Erik 
>> Castro de Lopo). The plan is to do as Robert Kern 
>> suggested: release the wrapper with the same license as 
>> the wrapped library (GPL here). 
>
> If you anticipate a reduction in usefulness from this,
> why not explain that and ask Erik for a release under the LGPL?
> Licenses are not a "given", especially when the number of 
> developers is small enough to feasibly agree on an 
> additional license.
I don't anticipate any reduction of usefulness. I was just wrong in my 
understanding that SRC was under the LGPL: Robert suggested to license a 
wrapper with the same license than the wrapped library, whatever the 
license is, and I myseful do not care about the licenses of my scikits.

Erik is a well known hacker in the audio linux programming scene, and he 
certainly is aware of the differences between GPL and LGPL. So although 
I agree asking for a relicensing for some part of the codes may be 
sometimes useful (and I did just that for my other audio related 
scikits, audiolab), I think asking to go from GPL to LGPL is quite rude 
(because it is GPL, Erik can dual license it to users who do not wish to 
open source their product, like Trolltech does with QT; with LGPL, not 
really)

cheers,

David
>
> Cheers,
> Alan Isaac
>
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>



From aisaac at american.edu  Tue Nov 20 23:18:30 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Tue, 20 Nov 2007 23:18:30 -0500
Subject: [SciPy-user] Adding pysamplerate to scikits ?
In-Reply-To: <4743975E.10705@ar.media.kyoto-u.ac.jp>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com><20071119142042.GA21903@localhost.cc.columbia.edu><4742546A.7090200@ar.media.kyoto-u.ac.jp>
	<Mahogany-0.67.0-1296-20071120-095159.00@american.edu><4743975E.10705@ar.media.kyoto-u.ac.jp>
Message-ID: <Mahogany-0.67.0-128-20071120-231830.00@american.edu>

On Wed, 21 Nov 2007, David Cournapeau apparently wrote:
> I think asking to go from GPL to LGPL is quite rude 
> (because it is GPL, Erik can dual license it to users who 
> do not wish to open source their product, like Trolltech 
> does with QT; with LGPL, not really)

Yes: if the GPL was chosen for that reason, I agree.
However many projects have chosen GPL by default,
without much though about licensing.  (This seems to
be getting rarer.)

Cheers,
Alan Isaac





From gael.varoquaux at normalesup.org  Wed Nov 21 02:45:22 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Wed, 21 Nov 2007 08:45:22 +0100
Subject: [SciPy-user] organizing data
In-Reply-To: <4742C339.8010308@cmla.ens-cachan.fr>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
	<20071119142042.GA21903@localhost.cc.columbia.edu>
	<4742546A.7090200@ar.media.kyoto-u.ac.jp>
	<4742C339.8010308@cmla.ens-cachan.fr>
Message-ID: <20071121074522.GA14595@clipper.ens.fr>

> My data are the variation of 11 parameters, with 18 values for each. so 
> you can see them as an hypercube of 11 dimension. My goal is the 
> possibility to plot with gnuplot in 2D (or 3D). So I have to class my 
> data by slices of 9 (or 8) parameters constant.
> That's my problem: it is not really an "hyper"matrix of data because i 
> did not discretized the space parameters equally for each parameters.
> So i think i have to class and then recursively cut into the mass so I 
> can have small arrays I can plot with the gnuplot.py

Hi Xavier,

Can you provide an example of your data?

If I understand properly, you have non-regular gridded data that you want
to interpolate on a regular grid (though only through a cut) to plot as
images. If I get it right, that does not sound an incredibly easy
problem. There are algorithmes to do this, but I don't think there are
wrapped in scipy. VTK but do this, but AFAIK, only in 3D.

But hopefully your problem is simpler, and if you show us the data we'll
understand it better.

Cheers,

Ga?l


From dmitrey.kroshko at scipy.org  Wed Nov 21 03:41:12 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Wed, 21 Nov 2007 10:41:12 +0200
Subject: [SciPy-user] a small OpenOpt example howto benchmark solvers
Message-ID: <4743EF28.4060707@scipy.org>

Hi all,
I have got a letter from Nils Wagner about interest of NLP solvers 
comparison, and maybe some of scipy-user mailist subscribers could be 
interested as well.

So if anyone is interested - here's an example of ALGENCAN, cobyla and 
lincher comparison:
http://openopt.blogspot.com/2007/11/example-of-oo-nlp-solvers-benchmark.html

Please let me note that usually ALGENCAN is much more better, and 
choosing correct stop criteria for your problem (gradtol, ftol, xtol 
etc) is very essential.

Regards, D


From boris at vonloesch.de  Wed Nov 21 03:54:38 2007
From: boris at vonloesch.de (Boris von Loesch)
Date: Wed, 21 Nov 2007 09:54:38 +0100
Subject: [SciPy-user] Sparse matrix advice
In-Reply-To: <2bc7a5a50711201530l25b3b813j673be7620c1e8154@mail.gmail.com>
References: <2bc7a5a50711201530l25b3b813j673be7620c1e8154@mail.gmail.com>
Message-ID: <41a2e2dc0711210054l476115d5l91054cfadb548f85@mail.gmail.com>

Hi Anand,

I use the COO-format for the creation of the matrix, because it could
be assembled quite fast. A second advantage is, that you need to
create the sparsity only the first time and then just create a vector
with the non-zero entries.
For calculations I convert it to CSR-Format, which is also quite fast.
With this method I can create a 255^2*255^2 sparse matrix with roughly
8 non-zero entries per row in about 0.25 sec. Creating the matrix in
LIL-format or DIC-format is a lot slower.

Best regards,
Boris

On Nov 21, 2007 12:30 AM, Anand Patil <anand.prabhakar.patil at gmail.com> wrote:
> Hi all,
>
> I'm looking for a 'row-ish' sparse matrix format that I can operate on
> (including making new nonzero elements) quickly in C++ or Pyrex. Could
> anyone recommend one?
>
> The LIL format looks nice because it can be expanded without any
> copying, but I'm worried that having to call PyList_GetItem (or
> whatever it is), inc/decreffing and converting from Python to C
> integers will take up a lot of time. PyTrilinos' Epetra looks pretty
> cool but I'm not sure how much distributed memory stuff I'll be doing
> and it would take me a long time to understand.
>
> I'm thinking that a good solution might be Pyrex wrappers for matrices
> represented as C++ vectors of rows, which are themselves vectors;
> would this actually be a good solution, and is anything like this out
> there already?
>
> I'm a total sparse linear algebra neophyte, so if I'm missing some
> boat entirely please help me get a clue.
>
> Many thanks in advance,
> Anand Patil
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From gael.varoquaux at normalesup.org  Wed Nov 21 05:16:50 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Wed, 21 Nov 2007 11:16:50 +0100
Subject: [SciPy-user] organizing data
In-Reply-To: <20071121074522.GA14595@clipper.ens.fr>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
	<20071119142042.GA21903@localhost.cc.columbia.edu>
	<4742546A.7090200@ar.media.kyoto-u.ac.jp>
	<4742C339.8010308@cmla.ens-cachan.fr>
	<20071121074522.GA14595@clipper.ens.fr>
Message-ID: <20071121101650.GE14595@clipper.ens.fr>

On Wed, Nov 21, 2007 at 08:45:22AM +0100, Gael Varoquaux wrote:
> If I understand properly, you have non-regular gridded data that you want
> to interpolate on a regular grid (though only through a cut) to plot as
> images. If I get it right, that does not sound an incredibly easy
> problem. There are algorithmes to do this, but I don't think there are
> wrapped in scipy. VTK but do this, but AFAIK, only in 3D.

I am stupid. Of course there are things in scipy !

First you can look at a Delaunay natural neighbor interpolation
(http://www.scipy.org/Cookbook/Matplotlib/Gridding_irregularly_spaced_data).
Unfortunately I think the algorithm in Scipy can only do this for 2D
data. Robert Kern will chime in, when he wakes up.

Second, you can use radial basis function for interpolation, see
http://www.scipy.org/Cookbook/RadialBasisFunctions
Unfortunately, you will have to rebuild scipy for this.


To everybody, is it possible to change the sandbox so that when scipy is
built, all the packages in the sandbox are built, but the build doesn't
complain if it fails for these (or you need a special switch to have it
complain). There are some really nice packages in the sandbox, but most
people cannot build them because they don't know how to build. IMHO it is
better to have the binary packages of scipy shipping with the fraction of
the sandbox that builds, and might not even pass tests, than to have
these packages simply not available for users who cannot build.

Cheers,

Ga?l


From j.reid at mail.cryst.bbk.ac.uk  Wed Nov 21 06:11:17 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Wed, 21 Nov 2007 11:11:17 +0000
Subject: [SciPy-user] Place data on grid to minimise cost function
Message-ID: <fi13q5$d0r$1@ger.gmane.org>

I have a problem for which I do not know any algorithm to solve it. I'm 
wondering if there is something in scipy that will help me. Apologies if 
there isn't and this is off-topic.

I have a set of N data points, which I want to place on a k x m grid. I 
have a cost function D(x1,x2) that measures how different 2 data points 
are. I want to minimise the sum of D over all neighbouring pairs of 
points on the grid (i.e. I want similar points to be neighbours on the 
grid).

A) Is there some established technique to do this?
B) Is it in scipy?

BTW, N = k.m

Thanks,
John.



From aisaac at american.edu  Wed Nov 21 09:47:35 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Wed, 21 Nov 2007 09:47:35 -0500
Subject: [SciPy-user] a small OpenOpt example howto benchmark solvers
In-Reply-To: <4743EF28.4060707@scipy.org>
References: <4743EF28.4060707@scipy.org>
Message-ID: <Mahogany-0.67.0-1456-20071121-094735.00@american.edu>

On Wed, 21 Nov 2007, dmitrey apparently wrote:
> http://openopt.blogspot.com/2007/11/example-of-oo-nlp-solvers-benchmark.html 

What problems did you run into with matplotlib?
Did you bring them to the matplotlib list?
(The developers are extremely responsive.)

Cheers,
Alan Isaac





From yfan at emory.edu  Wed Nov 21 09:50:29 2007
From: yfan at emory.edu (Ying Wai (Daniel) Fan)
Date: Wed, 21 Nov 2007 09:50:29 -0500
Subject: [SciPy-user] Any image deblurring and wavelets packages?
Message-ID: <474445B5.1070201@emory.edu>

Hi All,

I am a graduate student working on image deblurring. Is there any Python 
package doing that?
I am also interested in wavelets. The PyWavelets package hasn't been 
updated for a year. Is it still being developed? Are there other Python 
wavelets packages?

I want to persuade my adviser to switch from Matlab to Scipy. Without an 
image deblurring package/toolbox on Scipy, it is hard for him to switch. 
I am planning to rewrite some of his old codes in Scipy over the break 
and release them to the public. I just want to know if there is already 
some image deblurring Python codes out there?

Regards,
Daniel


From lev at columbia.edu  Wed Nov 21 10:04:48 2007
From: lev at columbia.edu (Lev Givon)
Date: Wed, 21 Nov 2007 10:04:48 -0500
Subject: [SciPy-user] Any image deblurring and wavelets packages?
In-Reply-To: <474445B5.1070201@emory.edu>
References: <474445B5.1070201@emory.edu>
Message-ID: <20071121150448.GA6392@localhost.cc.columbia.edu>

Received from Ying Wai (Daniel) Fan on Wed, Nov 21, 2007 at 09:50:29AM EST:
> Hi All,
> 
> I am a graduate student working on image deblurring. Is there any Python 
> package doing that?
> I am also interested in wavelets. The PyWavelets package hasn't been 
> updated for a year. Is it still being developed? Are there other Python 
> wavelets packages?
> 

The pywavelets code is still being actively developed:

http://wavelets.scipy.org

							L.G.


From xavier.barthelemy at cmla.ens-cachan.fr  Wed Nov 21 10:50:21 2007
From: xavier.barthelemy at cmla.ens-cachan.fr (Xavier Barthelemy)
Date: Wed, 21 Nov 2007 16:50:21 +0100
Subject: [SciPy-user] Place data on grid to minimise cost function
In-Reply-To: <fi13q5$d0r$1@ger.gmane.org>
References: <fi13q5$d0r$1@ger.gmane.org>
Message-ID: <474453BD.20706@cmla.ens-cachan.fr>

John Reid a ?crit :
> I have a problem for which I do not know any algorithm to solve it. I'm 
> wondering if there is something in scipy that will help me. Apologies if 
> there isn't and this is off-topic.
> 
> I have a set of N data points, which I want to place on a k x m grid. I 
> have a cost function D(x1,x2) that measures how different 2 data points 
> are. I want to minimise the sum of D over all neighbouring pairs of 
> points on the grid (i.e. I want similar points to be neighbours on the 
> grid).
> 
> A) Is there some established technique to do this?

yes. It is called (un)constraint optimization, or optimal control, or in 
physics, inverse problems. You can see Lions and al (he received a 
fields medal for that) for some biblio.

To do an optimal choice you will have to minimize or maximize the cost 
function with(out) constraint. you'll have to zeros the cost-function 
derivative respect to each parameters.
basically, you have 2 choices:
the sensibility method, where you manually (good!) or numerically (bouuu 
very bad!) calculate the frechet-derivative of the cost function, ie the 
derivative for each parameters respect to every else parameters for 
every single evaluation of your cost function. it means for k*m grid, 
you will need a (k*n)**2 matrix which represent the derivative of the 
cost function evaluated in one point.

the adjoint methods (very good, very modern, less calculus, stable 
numerically), where you evaluate the cost function derivative with the 
adjoint operator (mathematically calculated (better) or numerically 
calculated (not so acurate)) solved. The best method and the most 
stable, because it reimplace  an ill posed problem by a suite of well 
posed problem, in the Hadamard sense.

> B) Is it in scipy?

this one, I don't know :)

> 
> BTW, N = k.m
> 
> Thanks,
> John.
> 


cheers
Xavier



From gael.varoquaux at normalesup.org  Wed Nov 21 10:54:57 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Wed, 21 Nov 2007 16:54:57 +0100
Subject: [SciPy-user] organizing data
In-Reply-To: <474447DB.6030204@cmla.ens-cachan.fr>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
	<20071119142042.GA21903@localhost.cc.columbia.edu>
	<4742546A.7090200@ar.media.kyoto-u.ac.jp>
	<4742C339.8010308@cmla.ens-cachan.fr>
	<20071121074522.GA14595@clipper.ens.fr>
	<20071121101650.GE14595@clipper.ens.fr>
	<474447DB.6030204@cmla.ens-cachan.fr>
Message-ID: <20071121155457.GA2447@clipper.ens.fr>

Let's keep this on the mailing list, or else, we can switch to French
(not that I mind English).

On Wed, Nov 21, 2007 at 03:59:39PM +0100, Xavier Barthelemy wrote:
> So I think you did not understand my problem. It's a problem of 
> organizing and cutting by slice of rows in my data.


OK, you're telling me you already have interpolated your data, and you
have all the values you need (ie regular grids in the different
directions you are interested in), but you just need to sort them out.

> My data comports 18 values witch depends of 12 parameters.

> so I have my data like that:
> by rows:
> first the 12 parameters, then the 18 values.

> each rows represent a numeric experiment. I am doing parameters 
> exploration, so each varies independently. But the discretization of the 
> space parameters is not "regular": i have refined some for some values 
> of the others parameters.

If you want to do image plot, it will be more easier to have regular
data, that's where interpolation comes in. It looks to me like you want
to have an interpolation function f({P}) -> {V} where {P} is your set of
parameters and {V} your set of values. When you want to plot the cut
along a hyperplane HP, you simply choose a regular grid of this hyper
plane, and apply your interpolation function f on it. That's how I would
do it, if I understand your problem correctly.

> so my problem is really what I have (badly i guess) explained: I would 
> like to plot 2D (and 3D) graphs.
> Let's suppose that I have X ,Y and Z datas corresponding. knowing that, 
> how do you plot Y(X) with Z constant when you have a bunch of three 
> columns data?

In 1D that's really easy: if D is your [X, Y, Z] array, I would do

X = D[:, 0]
Y = D[:, 1]
Z = D[:, 2]

# Select all the data for Z = z0
x = X[Z==z0]
y = Y[Z==z0]

# Sort, just to make things prettier
y = y[argsort(x)]
x.sort()

plot(x, y)

> you 'll have to first sort by Z, and then by X. so when you'll plot 
> that, for each "X", sequentially sorted, you'll have the different "Y" 
> for each "Z" values. consequently you will have the number of different 
> Z values plots.

OK, you want to do this for all values of Z.

> And my problem now is the generalization with 12 parameters, let's name 
> them from A to J. How I'll do if I want to plot F(H)? the same, I will 
> sort by each of the 10 parameters and finally by H and I will have a 
> family of plots.

Yes, you can do this in a nice way using an array with fields and the
"order" argument to sort.

> but now I want to cut (slice) them to have each plot independently, and 
> i can plot them by the interfaced gnuplot.

You can always generalize the cutting method used in my example. If U, V,
W are parameters (similar to Z in the example above), you can define a
mask array:

mask = (U == u0) & (V == v0) & (W == w0)

# You dont really need x and y arrays, you could directly go to xy
x = X[mask]
y = Y[mask]
xy = empty(x.shape, dtype=dtype([('x','float'), ('y','float')]))
xy['x'] = x
xy['y'] = y

xy.sort(order=('x', 'y'))

then you have in xy['x'] and xy['y'] what you are interested in, if I
understand things right.

But I still think you need a regular grid, so either your data already
has that structure, and I don't understand why it has been "shuffled", or
it hasn't, and you'll need interpolating, so why bother sorting?
(selecting might be useful to reduce the amount of points).

HTH,

Ga?l


From josegomez at gmx.net  Wed Nov 21 10:59:57 2007
From: josegomez at gmx.net (Jose Luis Gomez Dans)
Date: Wed, 21 Nov 2007 16:59:57 +0100
Subject: [SciPy-user] Is scipy.io.numpyio.fread working?
Message-ID: <20071121155957.309730@gmx.net>

Hi,
I am trying to read binary data from a binary file. The file has a header, and several chunks of data in it. I want to seek some position, and read from there X data into an array. So, I have done the following:
fp = open( file )
fp.seek (header_position ) #My header position
a = scipy.io.numpyio.fread(fp, 2400*2400,'f',byteswap=1)

This bombs with a     
<type 'exceptions.TypeError'>: fread() takes no keyword arguments

I don't think this has to do with the fact that I seeked the file.
I am running scipy.__version__="0.6.0" from Andrew Straw's Ubuntu repository on Ubuntu i386.

Is there something wrong here?

Cheers!
Jose
-- 
GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
Alle Infos und kostenlose Anmeldung: http://www.gmx.net/de/go/freemail


From xavier.barthelemy at cmla.ens-cachan.fr  Wed Nov 21 11:24:10 2007
From: xavier.barthelemy at cmla.ens-cachan.fr (Xavier Barthelemy)
Date: Wed, 21 Nov 2007 17:24:10 +0100
Subject: [SciPy-user] organizing data
In-Reply-To: <20071121155457.GA2447@clipper.ens.fr>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
	<20071119142042.GA21903@localhost.cc.columbia.edu>
	<4742546A.7090200@ar.media.kyoto-u.ac.jp>
	<4742C339.8010308@cmla.ens-cachan.fr>
	<20071121074522.GA14595@clipper.ens.fr>
	<20071121101650.GE14595@clipper.ens.fr>
	<474447DB.6030204@cmla.ens-cachan.fr>
	<20071121155457.GA2447@clipper.ens.fr>
Message-ID: <47445BAA.7070406@cmla.ens-cachan.fr>

Gael Varoquaux a ?crit :
> Let's keep this on the mailing list, or else, we can switch to French
> (not that I mind English).

Sorry I made reply without looking at the email
> 
> On Wed, Nov 21, 2007 at 03:59:39PM +0100, Xavier Barthelemy wrote:
>> So I think you did not understand my problem. It's a problem of 
>> organizing and cutting by slice of rows in my data.
> 
> 
> OK, you're telling me you already have interpolated your data, and you
> have all the values you need (ie regular grids in the different
> directions you are interested in), but you just need to sort them out.
> 

no, I did not interpolate my data, because you cannot. I am looking for 
  sub-critical instabilities in binary instationary miscible 
compressible viscous fluids. so each numerical experiment is the result 
of (for some) days or hours of calculus in a high performance parallel 
computer.


>> My data comports 18 values witch depends of 12 parameters.
> 
>> so I have my data like that:
>> by rows:
>> first the 12 parameters, then the 18 values.
> 
>> each rows represent a numeric experiment. I am doing parameters 
>> exploration, so each varies independently. But the discretization of the 
>> space parameters is not "regular": i have refined some for some values 
>> of the others parameters.
> 
> If you want to do image plot, it will be more easier to have regular
> data, that's where interpolation comes in. It looks to me like you want
> to have an interpolation function f({P}) -> {V} where {P} is your set of
> parameters and {V} your set of values. When you want to plot the cut
> along a hyperplane HP, you simply choose a regular grid of this hyper
> plane, and apply your interpolation function f on it. That's how I would
> do it, if I understand your problem correctly.

Yes, you're right, i would do it like that, but I can't. Too much 
calculus. I am refining some zones in the space parameter independently.

> 
>> so my problem is really what I have (badly i guess) explained: I would 
>> like to plot 2D (and 3D) graphs.
>> Let's suppose that I have X ,Y and Z datas corresponding. knowing that, 
>> how do you plot Y(X) with Z constant when you have a bunch of three 
>> columns data?
> 
> In 1D that's really easy: if D is your [X, Y, Z] array, I would do
> 
> X = D[:, 0]
> Y = D[:, 1]
> Z = D[:, 2]
> 
> # Select all the data for Z = z0
> x = X[Z==z0]
> y = Y[Z==z0]
> 
> # Sort, just to make things prettier
> y = y[argsort(x)]
> x.sort()
> 
> plot(x, y)

i will try it. may be i have thought it too complicated.
> 
>> you 'll have to first sort by Z, and then by X. so when you'll plot 
>> that, for each "X", sequentially sorted, you'll have the different "Y" 
>> for each "Z" values. consequently you will have the number of different 
>> Z values plots.
> 
> OK, you want to do this for all values of Z.
> 
>> And my problem now is the generalization with 12 parameters, let's name 
>> them from A to J. How I'll do if I want to plot F(H)? the same, I will 
>> sort by each of the 10 parameters and finally by H and I will have a 
>> family of plots.
> 
> Yes, you can do this in a nice way using an array with fields and the
> "order" argument to sort.
> 
>> but now I want to cut (slice) them to have each plot independently, and 
>> i can plot them by the interfaced gnuplot.
> 
> You can always generalize the cutting method used in my example. If U, V,
> W are parameters (similar to Z in the example above), you can define a
> mask array:
> 
> mask = (U == u0) & (V == v0) & (W == w0)
> 
> # You dont really need x and y arrays, you could directly go to xy
> x = X[mask]
> y = Y[mask]
> xy = empty(x.shape, dtype=dtype([('x','float'), ('y','float')]))
> xy['x'] = x
> xy['y'] = y
> 
> xy.sort(order=('x', 'y'))
> 
> then you have in xy['x'] and xy['y'] what you are interested in, if I
> understand things right.

sounds good
> 
> But I still think you need a regular grid, so either your data already
> has that structure, and I don't understand why it has been "shuffled", or
> it hasn't, and you'll need interpolating, so why bother sorting?
> (selecting might be useful to reduce the amount of points).

they are "shuffled" because I did not saved a data-base with the logical 
links. so if I want the 7th column in function of the 5th, they are 
"shuffled"


thanks for your help, I'll try just now

cheers
Xavier



From josegomez at gmx.net  Wed Nov 21 11:25:43 2007
From: josegomez at gmx.net (Jose Luis Gomez Dans)
Date: Wed, 21 Nov 2007 17:25:43 +0100
Subject: [SciPy-user] Is scipy.io.numpyio.fread working?
In-Reply-To: <20071121155957.309730@gmx.net>
References: <20071121155957.309730@gmx.net>
Message-ID: <20071121162543.8250@gmx.net>

Hi,

-------- Original-Nachricht --------
> a = scipy.io.numpyio.fread(fp, 2400*2400,'f',byteswap=1)
> 
> This bombs with a     
> <type 'exceptions.TypeError'>: fread() takes no keyword arguments

The problem might be in the docstring. If I just do 
a = scipy.io.numpyio.fread(fp, 2400*2400,'f',byteswap=1).byteswap()
it works perfectly. However, the docstring appears to be from an older version of scipy where byteswapping was indeed part of fread.

Is it a bug?
Thanks,
Jose
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser


From josegomez at gmx.net  Wed Nov 21 11:25:43 2007
From: josegomez at gmx.net (Jose Luis Gomez Dans)
Date: Wed, 21 Nov 2007 17:25:43 +0100
Subject: [SciPy-user] Is scipy.io.numpyio.fread working?
In-Reply-To: <20071121155957.309730@gmx.net>
References: <20071121155957.309730@gmx.net>
Message-ID: <20071121162543.8250@gmx.net>

Hi,

-------- Original-Nachricht --------
> a = scipy.io.numpyio.fread(fp, 2400*2400,'f',byteswap=1)
> 
> This bombs with a     
> <type 'exceptions.TypeError'>: fread() takes no keyword arguments

The problem might be in the docstring. If I just do 
a = scipy.io.numpyio.fread(fp, 2400*2400,'f',byteswap=1).byteswap()
it works perfectly. However, the docstring appears to be from an older version of scipy where byteswapping was indeed part of fread.

Is it a bug?
Thanks,
Jose
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser


From gael.varoquaux at normalesup.org  Wed Nov 21 11:47:27 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Wed, 21 Nov 2007 17:47:27 +0100
Subject: [SciPy-user] organizing data
In-Reply-To: <47445BAA.7070406@cmla.ens-cachan.fr>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>
	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>
	<20071119142042.GA21903@localhost.cc.columbia.edu>
	<4742546A.7090200@ar.media.kyoto-u.ac.jp>
	<4742C339.8010308@cmla.ens-cachan.fr>
	<20071121074522.GA14595@clipper.ens.fr>
	<20071121101650.GE14595@clipper.ens.fr>
	<474447DB.6030204@cmla.ens-cachan.fr>
	<20071121155457.GA2447@clipper.ens.fr>
	<47445BAA.7070406@cmla.ens-cachan.fr>
Message-ID: <20071121164727.GE2447@clipper.ens.fr>

On Wed, Nov 21, 2007 at 05:24:10PM +0100, Xavier Barthelemy wrote:
> no, I did not interpolate my data, because you cannot. I am looking for 
>  sub-critical instabilities in binary instationary miscible 
> compressible viscous fluids. so each numerical experiment is the result 
> of (for some) days or hours of calculus in a high performance parallel 
> computer.

I am talking of interpolating using the results of these experiments, and
not by doing the calculations. Just think of you results as data points
and you need to do interpolation on these data points. If you don't have
regular gridded data, your plotting software will do interpolation (if it
knows how) to plot it with an image.

HTH,

Ga?l


From robert.kern at gmail.com  Wed Nov 21 13:37:01 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 21 Nov 2007 12:37:01 -0600
Subject: [SciPy-user] organizing data
In-Reply-To: <20071121101650.GE14595@clipper.ens.fr>
References: <4740FF05.4070302@ar.media.kyoto-u.ac.jp>	<c7009a550711182004p6ca49487k8c9e13178dab0a64@mail.gmail.com>	<20071119142042.GA21903@localhost.cc.columbia.edu>	<4742546A.7090200@ar.media.kyoto-u.ac.jp>	<4742C339.8010308@cmla.ens-cachan.fr>	<20071121074522.GA14595@clipper.ens.fr>
	<20071121101650.GE14595@clipper.ens.fr>
Message-ID: <47447ACD.7000608@gmail.com>

Gael Varoquaux wrote:
> On Wed, Nov 21, 2007 at 08:45:22AM +0100, Gael Varoquaux wrote:
>> If I understand properly, you have non-regular gridded data that you want
>> to interpolate on a regular grid (though only through a cut) to plot as
>> images. If I get it right, that does not sound an incredibly easy
>> problem. There are algorithmes to do this, but I don't think there are
>> wrapped in scipy. VTK but do this, but AFAIK, only in 3D.
> 
> I am stupid. Of course there are things in scipy !
> 
> First you can look at a Delaunay natural neighbor interpolation
> (http://www.scipy.org/Cookbook/Matplotlib/Gridding_irregularly_spaced_data).
> Unfortunately I think the algorithm in Scipy can only do this for 2D
> data. Robert Kern will chime in, when he wakes up.

You are correct.

> To everybody, is it possible to change the sandbox so that when scipy is
> built, all the packages in the sandbox are built, but the build doesn't
> complain if it fails for these (or you need a special switch to have it
> complain). There are some really nice packages in the sandbox, but most
> people cannot build them because they don't know how to build. IMHO it is
> better to have the binary packages of scipy shipping with the fraction of
> the sandbox that builds, and might not even pass tests, than to have
> these packages simply not available for users who cannot build.

Jarrod and I came to the conclusion that the sandbox packages will be moved out
of the sandbox. Packages that just need a home could go to scikits instead.
Packages intended for scipy itself can live under branches/ temporarily as a
standalone package until they are approved for inclusion. Maintainers who want
to build binaries of them can do so.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From nwagner at iam.uni-stuttgart.de  Wed Nov 21 14:04:07 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 21 Nov 2007 20:04:07 +0100
Subject: [SciPy-user] a small OpenOpt example howto benchmark solvers
In-Reply-To: <4743EF28.4060707@scipy.org>
References: <4743EF28.4060707@scipy.org>
Message-ID: <web-92353087@uni-stuttgart.de>

On Wed, 21 Nov 2007 10:41:12 +0200
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> Hi all,
> I have got a letter from Nils Wagner about interest of 
>NLP solvers 
> comparison, and maybe some of scipy-user mailist 
>subscribers could be 
> interested as well.
> 
> So if anyone is interested - here's an example of 
>ALGENCAN, cobyla and 
> lincher comparison:
> http://openopt.blogspot.com/2007/11/example-of-oo-nlp-solvers-benchmark.html
> 
> Please let me note that usually ALGENCAN is much more 
>better, and 
> choosing correct stop criteria for your problem 
>(gradtol, ftol, xtol 
> etc) is very essential.
> 
> Regards, D
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

  
Hi Dmitrey,

Do you have a reference for your benchmark example ?
I'm curious about it.

Nils


From dmitrey.kroshko at scipy.org  Wed Nov 21 14:11:32 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Wed, 21 Nov 2007 21:11:32 +0200
Subject: [SciPy-user] a small OpenOpt example howto benchmark solvers
In-Reply-To: <web-92353087@uni-stuttgart.de>
References: <4743EF28.4060707@scipy.org> <web-92353087@uni-stuttgart.de>
Message-ID: <474482E4.8060105@scipy.org>

Hi Nils,
I suppose I have provided all required URLs.
Let me mention the ones once again:

blog report about results
http://openopt.blogspot.com/2007/11/example-of-oo-nlp-solvers-benchmark.html
and this one contains URLs to code (7th line from top, "the example"):
http://projects.scipy.org/scipy/scikits/browser/trunk/openopt/scikits/openopt/examples/nlp_bench_1.py

Alan G Isaac wrote:
> What problems did you run into with matplotlib?
> Did you bring them to the matplotlib list?
> (The developers are extremely responsive.)
Thank you, I'll take it into account.
Regards, D.

Nils Wagner wrote:
> On Wed, 21 Nov 2007 10:41:12 +0200
>   dmitrey <dmitrey.kroshko at scipy.org> wrote:
>   
>> Hi all,
>> I have got a letter from Nils Wagner about interest of 
>> NLP solvers 
>> comparison, and maybe some of scipy-user mailist 
>> subscribers could be 
>> interested as well.
>>
>> So if anyone is interested - here's an example of 
>> ALGENCAN, cobyla and 
>> lincher comparison:
>> http://openopt.blogspot.com/2007/11/example-of-oo-nlp-solvers-benchmark.html
>>
>> Please let me note that usually ALGENCAN is much more 
>> better, and 
>> choosing correct stop criteria for your problem 
>> (gradtol, ftol, xtol 
>> etc) is very essential.
>>
>> Regards, D
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>     
>
>   
> Hi Dmitrey,
>
> Do you have a reference for your benchmark example ?
> I'm curious about it.
>
> Nils
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From nwagner at iam.uni-stuttgart.de  Wed Nov 21 14:28:08 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 21 Nov 2007 20:28:08 +0100
Subject: [SciPy-user] a small OpenOpt example howto benchmark solvers
In-Reply-To: <474482E4.8060105@scipy.org>
References: <4743EF28.4060707@scipy.org> <web-92353087@uni-stuttgart.de>
	<474482E4.8060105@scipy.org>
Message-ID: <web-92353658@uni-stuttgart.de>

On Wed, 21 Nov 2007 21:11:32 +0200
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> Hi Nils,
> I suppose I have provided all required URLs.
> Let me mention the ones once again:
> 
> blog report about results
> http://openopt.blogspot.com/2007/11/example-of-oo-nlp-solvers-benchmark.html
> and this one contains URLs to code (7th line from top, 
>"the example"):
> http://projects.scipy.org/scipy/scikits/browser/trunk/openopt/scikits/openopt/examples/nlp_bench_1.py
> 
  
Sorry for the confusion. I was talking about a reference 
from the literature.

Nils


From dmitrey.kroshko at scipy.org  Wed Nov 21 14:57:04 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Wed, 21 Nov 2007 21:57:04 +0200
Subject: [SciPy-user] a small OpenOpt example howto benchmark solvers
In-Reply-To: <web-92353658@uni-stuttgart.de>
References: <4743EF28.4060707@scipy.org>
	<web-92353087@uni-stuttgart.de>	<474482E4.8060105@scipy.org>
	<web-92353658@uni-stuttgart.de>
Message-ID: <47448D90.9090008@scipy.org>

Nils Wagner wrote:
> On Wed, 21 Nov 2007 21:11:32 +0200
>   dmitrey <dmitrey.kroshko at scipy.org> wrote:
>   
>> Hi Nils,
>> I suppose I have provided all required URLs.
>> Let me mention the ones once again:
>>
>> blog report about results
>> http://openopt.blogspot.com/2007/11/example-of-oo-nlp-solvers-benchmark.html
>> and this one contains URLs to code (7th line from top, 
>> "the example"):
>> http://projects.scipy.org/scipy/scikits/browser/trunk/openopt/scikits/openopt/examples/nlp_bench_1.py
>>
>>     
>   
> Sorry for the confusion. I was talking about a reference 
> from the literature.
>
> Nils
>   
That example is not from literature, it's just almost 1st that I decided 
to code and bench (looks very similar to nlp_2.py from 
/openopt/examples/, but objfunc is (x-M)**1.5, not **2, to prevent 100% 
pure quadratic func).
Almost all examples from literature available for me are too small to be 
taken into openopt time and cputime comparison, because (as I have 
mentioned you in a letter) openopt requires some time for preparation 
prob struct before and after calculations; also, stop criteria mean too 
much for so small problems.
As you see from the example, ALGENCAN takes much more f,c,h evals (for 
*this* example) time is almost same. It's because the cause I have 
mentioned (it's espesially valid for small problems), + ALGENCAN is 
fortran-written (as well as cobyla), while lincher is 100% Python + 
cvxopt QP solver, that is also mostly Python-written.
Regards, D.


From robert.kern at gmail.com  Wed Nov 21 15:17:26 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 21 Nov 2007 14:17:26 -0600
Subject: [SciPy-user] help on array manipulation
In-Reply-To: <loom.20071120T163815-497@post.gmane.org>
References: <loom.20071120T163815-497@post.gmane.org>
Message-ID: <47449256.4070407@gmail.com>

Tim Michelsen wrote:
> Hello,
> 
> I have the following array:
> 
> In [240]: z
> Out[240]:
> array([[ 15.,   3.,   7.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
>           0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
>           0.,   0.,   0.],
>        [ 16.,   3.,   7.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
>           0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
>           0.,   0.,   0.],
>        [ 17.,   3.,   7.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
>           0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,   0.,
>           0.,   0.,   0.]])
> 
> How do I get it be become like:
> 
> 15 3
> 15 7
> 15 0
> ...
> 16 3
> 16 7
> ...
> 17 3
> 
> 
> Is there a special function I can use for this or do I need to build various for
> loops to achive this?

import numpy

keys = z[:,0]
values = z[:,1:]
n = values.shape[-1]
z2 = numpy.column_stack([numpy.repeat(keys, n), values.flat])

> I mean this example has few data. But what if the amount increases?
> 
> I am stuck here.
> 
> I'd be very glad if anyone could point be to a good tutorial for array
> manipulation. Scipy Cookbook is not enough.
> Well, I also found :
> http://numpy.scipy.org/numpydoc
> and
> http://www.penzilla.net/tutorials/python/numeric

This helps:
http://www.scipy.org/Numpy_Example_List_With_Doc

but honestly, it mostly comes with practice.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From anand.prabhakar.patil at gmail.com  Wed Nov 21 21:58:38 2007
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Wed, 21 Nov 2007 18:58:38 -0800
Subject: [SciPy-user] SciPy-user Digest, Vol 51, Issue 45
In-Reply-To: <mailman.19354.1195643545.2308.scipy-user@scipy.org>
References: <mailman.19354.1195643545.2308.scipy-user@scipy.org>
Message-ID: <2bc7a5a50711211858y76f766d1i4816cfe3b45f2e22@mail.gmail.com>

>From: "Boris von Loesch" <boris at vonloesch.de>
> Subject: Re: [SciPy-user] Sparse matrix advice
>
> Hi Anand,
>
> I use the COO-format for the creation of the matrix, because it could
> be assembled quite fast. A second advantage is, that you need to
> create the sparsity only the first time and then just create a vector
> with the non-zero entries.
> For calculations I convert it to CSR-Format, which is also quite fast.
> With this method I can create a 255^2*255^2 sparse matrix with roughly
> 8 non-zero entries per row in about 0.25 sec. Creating the matrix in
> LIL-format or DIC-format is a lot slower.
>
> Best regards,
> Boris

Hi Boris,

Thanks for the advice, I think I can get where I'm trying to go now. A
couple of additional questions: Is there a triangular version (DTRMM)
of csrmucsr out there somewhere? How about a CSR-CSR triangular solve?

Thanks,
Anand


From nwagner at iam.uni-stuttgart.de  Thu Nov 22 07:36:57 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 22 Nov 2007 13:36:57 +0100
Subject: [SciPy-user] Read an array from a file using io.read_array
Message-ID: <web-92386922@uni-stuttgart.de>

Hi all,

Is it possible to use io.read_array for the following 
task.

I have an ascii file realeig.pmat.
The essential information in this file is a table with 4 
columns and several lines corresponding to the number of 
computed eigenvalues.
The task is to read that information from a file.
Empty lines and lines beginning with "!", "$" should be 
skipped.
How can I handle the delimiter "&" which is used to 
continue lines ?

Any pointer would be appreciated.

Nils
-------------- next part --------------
A non-text attachment was scrubbed...
Name: realeig.pmat.gz
Type: application/x-gzip
Size: 1057 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071122/dcd55fe4/attachment.bin>

From ptenconi at eco.uninsubria.it  Thu Nov 22 09:24:42 2007
From: ptenconi at eco.uninsubria.it (Paolo Tenconi)
Date: Thu, 22 Nov 2007 15:24:42 +0100
Subject: [SciPy-user] Numpy matrix and blitz Matrix
Message-ID: <op.t166bgg2qiheo8@computer-di-paolo-tenconi.local>

Hi Everyone,

*) numpy arrays are automatically converted to blitz arrays. That's fine.

*) I need to work with blitz Matrix objects, but I noticed that
    numpy matrix objects are not converted to them and I get a
    compilation error too

3) Does someone on the list succeeded doing that?
    Or is there a workaround to create quickly a blitz matrix object
    from a numpy 2D array or numpy-matrix without wasting time
    in casting operations?

It would be nice having an automatic translation of numpy.matrix to
blitz Matrix as matrices are used frequently in scientific computing.

Any suggestion or help would be very glad.
Paolo


#------------------  EXAMPLE CODE -----------------------------------

import numpy
import scipy

#Dot product with arrays (it works)
x=numpy.array([[1,2,],[3,4]])
y=numpy.zeros((2,2))
scipy.weave.inline("""y=x*x;""",['x','y'],type_converters=scipy.weave.converters.blitz,compiler='gcc',force=1)

#Matrix multiplication with matrix (gives compilation error)
X=numpy.matlib.mat('1 2; 3 4')
Y=numpy.matlib.zeros((2,2))
scipy.weave.inline("""Y=X*X;""",['X','Y'],type_converters=scipy.weave.converters.blitz,compiler='gcc')

#--------------------------------------------------------------------


From aisaac at american.edu  Thu Nov 22 11:04:04 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Thu, 22 Nov 2007 11:04:04 -0500
Subject: [SciPy-user] Read an array from a file using io.read_array
In-Reply-To: <web-92386922@uni-stuttgart.de>
References: <web-92386922@uni-stuttgart.de>
Message-ID: <Mahogany-0.67.0-288-20071122-110404.00@american.edu>

On Thu, 22 Nov 2007, Nils Wagner apparently wrote:
> Is it possible to use io.read_array for the following 
> task. 

I think you want numpy.loadtxt

Cheers,
Alan Isaac





From nwagner at iam.uni-stuttgart.de  Thu Nov 22 12:00:42 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 22 Nov 2007 18:00:42 +0100
Subject: [SciPy-user] Read an array from a file using io.read_array
In-Reply-To: <Mahogany-0.67.0-288-20071122-110404.00@american.edu>
References: <web-92386922@uni-stuttgart.de>
	<Mahogany-0.67.0-288-20071122-110404.00@american.edu>
Message-ID: <web-92404974@uni-stuttgart.de>

On Thu, 22 Nov 2007 11:04:04 -0500
  Alan G Isaac <aisaac at american.edu> wrote:
> On Thu, 22 Nov 2007, Nils Wagner apparently wrote:
>> Is it possible to use io.read_array for the following 
>> task. 
> 
> I think you want numpy.loadtxt
> 
> Cheers,
> Alan Isaac
> 
> 
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

loadtxt(fname, dtype=<type 'float'>, comments='#', 
delimiter=None, converters=None, skiprows=0, usecols=None, 
unpack=False)

Can I use more than one character to filter out comment 
lines starting with "!" or "$" ?

e.g. comments='!$'


Nils


From rshepard at appl-ecosys.com  Thu Nov 22 13:04:43 2007
From: rshepard at appl-ecosys.com (Rich Shepard)
Date: Thu, 22 Nov 2007 10:04:43 -0800 (PST)
Subject: [SciPy-user] Web Site's Down
Message-ID: <Pine.LNX.4.64.0711221003191.13934@salmo.appl-ecosys.com>

   Did the server for the scipy.org web site take off for the Thanksgiving
holiday? Firefox keeps reporting that the server cannot be found.

Rich

-- 
Richard B. Shepard, Ph.D.               |  Integrity            Credibility
Applied Ecosystem Services, Inc.        |            Innovation
<http://www.appl-ecosys.com>     Voice: 503-667-4517      Fax: 503-667-8863


From gael.varoquaux at normalesup.org  Thu Nov 22 13:07:10 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Thu, 22 Nov 2007 19:07:10 +0100
Subject: [SciPy-user] Web Site's Down
In-Reply-To: <Pine.LNX.4.64.0711221003191.13934@salmo.appl-ecosys.com>
References: <Pine.LNX.4.64.0711221003191.13934@salmo.appl-ecosys.com>
Message-ID: <20071122180710.GI4775@clipper.ens.fr>

On Thu, Nov 22, 2007 at 10:04:43AM -0800, Rich Shepard wrote:
>    Did the server for the scipy.org web site take off for the Thanksgiving
> holiday? Firefox keeps reporting that the server cannot be found.

Its fine here ! No thanksgivinf in France.

Cheers,

Ga?l


From rshepard at appl-ecosys.com  Thu Nov 22 13:10:58 2007
From: rshepard at appl-ecosys.com (Rich Shepard)
Date: Thu, 22 Nov 2007 10:10:58 -0800 (PST)
Subject: [SciPy-user] Web Site's Down
In-Reply-To: <20071122180710.GI4775@clipper.ens.fr>
References: <Pine.LNX.4.64.0711221003191.13934@salmo.appl-ecosys.com>
	<20071122180710.GI4775@clipper.ens.fr>
Message-ID: <Pine.LNX.4.64.0711221009390.13934@salmo.appl-ecosys.com>

On Thu, 22 Nov 2007, Gael Varoquaux wrote:

> Its fine here ! No thanksgivinf in France.

Ga?l,

   It came back up here just a few minutes ago.

   And you might have a thanksgiving if the rail and energy workers stop
striking and go back to work. :-)

Rich

-- 
Richard B. Shepard, Ph.D.               |  Integrity            Credibility
Applied Ecosystem Services, Inc.        |            Innovation
<http://www.appl-ecosys.com>     Voice: 503-667-4517      Fax: 503-667-8863

From vagabondaero at gmail.com  Thu Nov 22 13:23:36 2007
From: vagabondaero at gmail.com (Vagabond_Aero)
Date: Thu, 22 Nov 2007 10:23:36 -0800
Subject: [SciPy-user] Web Site's Down
In-Reply-To: <Pine.LNX.4.64.0711221009390.13934@salmo.appl-ecosys.com>
References: <Pine.LNX.4.64.0711221003191.13934@salmo.appl-ecosys.com>
	<20071122180710.GI4775@clipper.ens.fr>
	<Pine.LNX.4.64.0711221009390.13934@salmo.appl-ecosys.com>
Message-ID: <8b5ec91a0711221023hb18d50ck5a48bb4d62cd2347@mail.gmail.com>

It popped right up for me

On Nov 22, 2007 10:10 AM, Rich Shepard <rshepard at appl-ecosys.com> wrote:
> On Thu, 22 Nov 2007, Gael Varoquaux wrote:
>
> > Its fine here ! No thanksgivinf in France.
>
> Ga?l,
>
>    It came back up here just a few minutes ago.
>
>    And you might have a thanksgiving if the rail and energy workers stop
> striking and go back to work. :-)
>
>
> Rich
>
> --
> Richard B. Shepard, Ph.D.               |  Integrity            Credibility
> Applied Ecosystem Services, Inc.        |            Innovation
> <http://www.appl-ecosys.com>     Voice: 503-667-4517      Fax: 503-667-8863
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


From fperez.net at gmail.com  Thu Nov 22 16:18:47 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Thu, 22 Nov 2007 14:18:47 -0700
Subject: [SciPy-user] New example
In-Reply-To: <20071029210726.GE30790@mentat.za.net>
References: <ce557a360710251321j7afec1d8y6346db50e5784f91@mail.gmail.com>
	<20071025211521.GA15301@mentat.za.net>
	<20071026064019.GD9025@clipper.ens.fr>
	<20071029210726.GE30790@mentat.za.net>
Message-ID: <db6b5ecc0711221318m5b83af92l300800cf54cfcfbd@mail.gmail.com>

On Oct 29, 2007 2:07 PM, Stefan van der Walt <stefan at sun.ac.za> wrote:
> Hi all,
>
> Here is the code for doing an elementary finite difference heat-flow
> simulation.  This is really just a repeated convolution, with certain
> values resetting after each time-step (sources or sinks).  The code
> was written for Numeric, but it seems to run under NumPy without
> problems.
>
> Regards
> St?fan
>
>
> On Fri, Oct 26, 2007 at 08:40:19AM +0200, Gael Varoquaux wrote:
> > Anne and Stefan, I think these examples are very useful, and definitely
> > should be on the scipy.org website.

Would you all mind/want to have them contributed to the workbook of
the 'python for science' workshop that John Hunter and I have taught
(and will teach again soon)?

http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/py4science/workbook/

Obviously contributor credit would be given for all material, and this
is available already to anyone wanting to use this for teaching (I
need to update the credits page to list Stefan already for the FFT
examples, let me know of anyone else I've missed).

For a long time we've talked about sharing open teaching materials for
scientific computing, and this is already a little start.  The current
workbook with solutions clocks in at ~57 pages, and over the next few
days we'll try to add more.

Cheers,

f


From peridot.faceted at gmail.com  Thu Nov 22 17:16:15 2007
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Thu, 22 Nov 2007 17:16:15 -0500
Subject: [SciPy-user] New example
In-Reply-To: <db6b5ecc0711221318m5b83af92l300800cf54cfcfbd@mail.gmail.com>
References: <ce557a360710251321j7afec1d8y6346db50e5784f91@mail.gmail.com>
	<20071025211521.GA15301@mentat.za.net>
	<20071026064019.GD9025@clipper.ens.fr>
	<20071029210726.GE30790@mentat.za.net>
	<db6b5ecc0711221318m5b83af92l300800cf54cfcfbd@mail.gmail.com>
Message-ID: <ce557a360711221416x5092f54ayde4afa67227a9557@mail.gmail.com>

On 22/11/2007, Fernando Perez <fperez.net at gmail.com> wrote:

> Would you all mind/want to have them contributed to the workbook of
> the 'python for science' workshop that John Hunter and I have taught
> (and will teach again soon)?

That would be great!

Anne


From aisaac at american.edu  Thu Nov 22 23:14:08 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Thu, 22 Nov 2007 23:14:08 -0500
Subject: [SciPy-user] Read an array from a file using io.read_array
In-Reply-To: <web-92404974@uni-stuttgart.de>
References: <web-92386922@uni-stuttgart.de><Mahogany-0.67.0-288-20071122-110404.00@american.edu><web-92404974@uni-stuttgart.de>
Message-ID: <Mahogany-0.67.0-288-20071122-231408.00@american.edu>

> On Thu, 22 Nov 2007 11:04:04 -0500 Alan wrote:
>> I think you want numpy.loadtxt 


On Thu, 22 Nov 2007, Nils Wagner apparently wrote:
> Can I use more than one character to filter out comment 
> lines starting with "!" or "$" ? 
> e.g. comments='!$'


As long as all comments start in the same way and these two 
characters always start comments sequentially, no problem.
But I think you are saying there are two **different** 
comment characters?  Then you will have to filter comments 
yourself.

Cheers,
Alan Isaac




From ramercer at gmail.com  Thu Nov 22 23:54:26 2007
From: ramercer at gmail.com (Adam Mercer)
Date: Thu, 22 Nov 2007 23:54:26 -0500
Subject: [SciPy-user] Building with a non-standard gfortran
In-Reply-To: <799406d60711200603t8e451cah2e13add2bd794a11@mail.gmail.com>
References: <799406d60711200603t8e451cah2e13add2bd794a11@mail.gmail.com>
Message-ID: <799406d60711222054x696c6908tfcc560d498e043b8@mail.gmail.com>

On 20/11/2007, Adam Mercer <ramercer at gmail.com> wrote:
> Hi
>
> Using MacPorts, the gfortran compiler, from gcc-4.2.x, is installed as
> gfortran-mp-4.2.  I have been trying to build scipy with this compiler
> and found that that numpy I need to build and install with
>
> $ python setup.py config_fc --fcompiler gnu95 \
>     --f77exec /opt/local/bin/gfortran-mp-4.2 \
>     --f90exec /opt/local/bin/gfortran-mp-4.2 build
> $ python setup.py install --prefix=${NUMPY_LOCATION}
>
> however using the above for scipy fails on the install phase with it
> being unable to locate a fortran compiler.  I therefore need to
> install scipy with
>
> $ python setup.py config_fc --fcompiler gnu95 \
>     --f77exec /opt/local/bin/gfortran-mp-4.2 \
>     --f90exec /opt/local/bin/gfortran-mp-4.2 install --prefix=${SCIPY_LOCATION}
>
> Is this expected, as I would have expected install to inherit the
> options that where used during the build phase?
>
> Cheers
>
> Adam
>
> PS: I'm using numpy-1.0.4 and scipy-0.6.0

Anyone?


From cookedm at physics.mcmaster.ca  Fri Nov 23 02:31:10 2007
From: cookedm at physics.mcmaster.ca (David M. Cooke)
Date: Fri, 23 Nov 2007 02:31:10 -0500
Subject: [SciPy-user] Building with a non-standard gfortran
In-Reply-To: <799406d60711200603t8e451cah2e13add2bd794a11@mail.gmail.com>
References: <799406d60711200603t8e451cah2e13add2bd794a11@mail.gmail.com>
Message-ID: <4BD1DAD9-E9B8-4AE2-93E5-FE46B146830D@physics.mcmaster.ca>

On Nov 20, 2007, at 09:03 , Adam Mercer wrote:

> Hi
>
> Using MacPorts, the gfortran compiler, from gcc-4.2.x, is installed as
> gfortran-mp-4.2.  I have been trying to build scipy with this compiler
> and found that that numpy I need to build and install with
>
> $ python setup.py config_fc --fcompiler gnu95 \
>    --f77exec /opt/local/bin/gfortran-mp-4.2 \
>    --f90exec /opt/local/bin/gfortran-mp-4.2 build
> $ python setup.py install --prefix=${NUMPY_LOCATION}
>
> however using the above for scipy fails on the install phase with it
> being unable to locate a fortran compiler.  I therefore need to
> install scipy with
>
> $ python setup.py config_fc --fcompiler gnu95 \
>    --f77exec /opt/local/bin/gfortran-mp-4.2 \
>    --f90exec /opt/local/bin/gfortran-mp-4.2 install --prefix=$ 
> {SCIPY_LOCATION}
>
> Is this expected, as I would have expected install to inherit the
> options that where used during the build phase?

The options used in the build phase aren't stored, so, yes, they need  
to be respecified. For numpy it doesn't matter, as the Fortran  
compiler isn't used. Since we use distutils, you can set the config_fc  
options in a file (either globally for the user in a  
~/.pydistutils.cfg, or per-package in a setup.cfg next to the setup.py  
you're running) by adding the following section to the appropiate file

[config_fc]
fcompiler=gfortran
f77exec=gfortran-mp-4.2
f90exec=gfortran-mp-4.2

-- 
|>|\/|<
/------------------------------------------------------------------\
|David M. Cooke              http://arbutus.physics.mcmaster.ca/dmc/
|cookedm at physics.mcmaster.ca


From dbm.sun at gmail.com  Fri Nov 23 12:18:03 2007
From: dbm.sun at gmail.com (Yutec Sun)
Date: Fri, 23 Nov 2007 12:18:03 -0500
Subject: [SciPy-user] scipy 0.6 installation problem
Message-ID: <11001d3e0711230918q1ab5e650s53c98fe8613420dd@mail.gmail.com>

Hello,
I'd like to ask your favor for helping me install scipy 0.6.
I have a problem importing scipy. In fact, importing doesn't show an
error message, but scipy.test produces failures.
My OS is Ubuntu 7.10 (Gutsy Gibbon) amd64. Following the installation
instruction, I attach the messages for further information.
Please understand that I'm new to linux system and scipy, and let me
know how to successfully install scipy step by step.
I thank you for your help.

Best regards,

Yutec


--------------------------------------------------------------------------------------------------------------
1. Error message
--------------------------------------------------------------------------------------------------------------
yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy/distutils$ python
Python 2.5.1 (r251:54863, Oct  5 2007, 13:50:07)
[GCC 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import scipy
>>> scipy.test(level=1)
  Found 9/9 tests for scipy.cluster.vq
  Found 18/18 tests for scipy.fftpack.basic
  Found 4/4 tests for scipy.fftpack.helper
  Found 20/20 tests for scipy.fftpack.pseudo_diffs
  Found 1/1 tests for scipy.integrate
  Found 10/10 tests for scipy.integrate.quadpack
  Found 3/3 tests for scipy.integrate.quadrature
  Found 6/6 tests for scipy.interpolate
  Found 6/6 tests for scipy.interpolate.fitpack
  Found 4/4 tests for scipy.io.array_import
  Found 28/28 tests for scipy.io.mio
  Found 13/13 tests for scipy.io.mmio
  Found 5/5 tests for scipy.io.npfile
  Found 4/4 tests for scipy.io.recaster
  Found 16/16 tests for scipy.lib.blas
  Found 128/128 tests for scipy.lib.blas.fblas
  Found 42/42 tests for scipy.lib.lapack
  Found 41/41 tests for scipy.linalg.basic
<module 'scipy.linalg.fblas' from
'/usr/lib/python2.5/site-packages/scipy/linalg/fblas.so'>
  Found 16/16 tests for scipy.linalg.blas
  Found 72/72 tests for scipy.linalg.decomp
  Found 128/128 tests for scipy.linalg.fblas
  Found 6/6 tests for scipy.linalg.iterative
  Found 4/4 tests for scipy.linalg.lapack
  Found 7/7 tests for scipy.linalg.matfuncs
  Found 9/9 tests for scipy.linsolve.umfpack
  Found 2/2 tests for scipy.maxentropy
  Found 3/3 tests for scipy.misc.pilutil
  Found 399/399 tests for scipy.ndimage
  Found 5/5 tests for scipy.odr
  Found 8/8 tests for scipy.optimize
  Found 1/1 tests for scipy.optimize.cobyla
  Found 10/10 tests for scipy.optimize.nonlin
  Found 4/4 tests for scipy.optimize.zeros
  Found 5/5 tests for scipy.signal.signaltools
  Found 4/4 tests for scipy.signal.wavelets
  Found 152/152 tests for scipy.sparse
  Found 342/342 tests for scipy.special.basic
  Found 3/3 tests for scipy.special.spfun_stats
  Found 107/107 tests for scipy.stats
  Found 73/73 tests for scipy.stats.distributions
  Found 10/10 tests for scipy.stats.morestats
  Found 0/0 tests for __main__
/usr/lib/python2.5/site-packages/scipy/cluster/vq.py:477: UserWarning:
One of the clusters is empty. Re-run kmean with a different
initialization.
  warnings.warn("One of the clusters is empty. "
...exception raised as expected: One of the clusters is empty. Re-run
kmean with a different initialization.
................................................Residual: 1.05006987327e-07
..................../usr/lib/python2.5/site-packages/scipy/interpolate/fitpack2.py:458:
UserWarning:
The coefficients of the spline returned have been computed as the
minimal norm least-squares solution of a (numerically) rank deficient
system (deficiency=7). If deficiency is large, the results may be
inaccurate. Deficiency may strongly depend on the value of eps.
  warnings.warn(message)
......
Don't worry about a warning regarding the number of bytes read.
Warning: 1000000 bytes requested, 20 bytes read.
.........................................................................caxpy:n=4
..caxpy:n=3
....ccopy:n=4
..ccopy:n=3
.............cscal:n=4
....cswap:n=4
..cswap:n=3
.....daxpy:n=4
..daxpy:n=3
....dcopy:n=4
..dcopy:n=3
.............dscal:n=4
....dswap:n=4
..dswap:n=3
.....saxpy:n=4
..saxpy:n=3
....scopy:n=4
..scopy:n=3
.............sscal:n=4
....sswap:n=4
..sswap:n=3
.....zaxpy:n=4
..zaxpy:n=3
....zcopy:n=4
..zcopy:n=3
.............zscal:n=4
....zswap:n=4
..zswap:n=3
................................................................................................................................................................................caxpy:n=4
..caxpy:n=3
....ccopy:n=4
..ccopy:n=3
.............cscal:n=4
....cswap:n=4
..cswap:n=3
.....daxpy:n=4
..daxpy:n=3
....dcopy:n=4
..dcopy:n=3
.............dscal:n=4
....dswap:n=4
..dswap:n=3
.....saxpy:n=4
..saxpy:n=3
....scopy:n=4
..scopy:n=3
.............sscal:n=4
....sswap:n=4
..sswap:n=3
.....zaxpy:n=4
..zaxpy:n=3
....zcopy:n=4
..zcopy:n=3
.............zscal:n=4
....zswap:n=4
..zswap:n=3
.............Result may be inaccurate, approximate err = 1.56163357201e-08
...Result may be inaccurate, approximate err = 1.25221600096e-10
......Use minimum degree ordering on A'+A.
..Use minimum degree ordering on A'+A.
...Use minimum degree ordering on A'+A.
............................................................................................................./usr/lib/python2.5/site-packages/scipy/ndimage/interpolation.py:41:
UserWarning: Mode "reflect" may yield incorrect results on boundaries.
Please use "mirror" instead.
  warnings.warn('Mode "reflect" may yield incorrect results on '
........................................................................................................................................................................................................................................................................................................F..F.........................................................Use
minimum degree ordering on A'+A.
.....................................Use minimum degree ordering on A'+A.
.....................................Use minimum degree ordering on A'+A.
................................Use minimum degree ordering on A'+A.
....................................................................................................................................................................................................................................................................................................................................................0.2
0.2
0.2
......0.2
..0.2
0.2
0.2
0.2
0.2
.........................................................................................................................................................................................................Ties
preclude use of exact statistic.
..Ties preclude use of exact statistic.
......
======================================================================
FAIL: test_explicit (scipy.tests.test_odr.test_odr)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
line 50, in test_explicit
    -8.7849712165253724e-02]),
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 100.0%)
 x: array([  1.26462971e+03,  -5.42545890e+01,  -8.64250389e-02])
 y: array([  1.26465481e+03,  -5.40184100e+01,  -8.78497122e-02])

======================================================================
FAIL: test_multi (scipy.tests.test_odr.test_odr)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
line 191, in test_multi
    0.5101147161764654,  0.5173902330489161]),
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 100.0%)
 x: array([ 4.31272063,  2.44289312,  7.76215871,  0.55995622,  0.46423343])
 y: array([ 4.37998803,  2.43330576,  8.00288459,  0.51011472,  0.51739023])

----------------------------------------------------------------------
Ran 1728 tests in 3.660s

FAILED (failures=2)
<unittest._TextTestResult run=1728 errors=0 failures=2>



--------------------------------------------------------------------------------------------------------------
2. System check
--------------------------------------------------------------------------------------------------------------
yutec at yutec-laptop:~$ python -c 'import os,sys;print os.name,sys.platform'
posix linux2

yutec at yutec-laptop:~$ uname -a
Linux yutec-laptop 2.6.22-14-generic #1 SMP Sun Oct 14 21:45:15 GMT
2007 x86_64 GNU/Linux

yutec at yutec-laptop:~$ gcc -v
Using built-in specs.
Target: x86_64-linux-gnu
Configured with: ../src/configure -v
--enable-languages=c,c++,fortran,objc,obj-c++,treelang --prefix=/usr
--enable-shared --with-system-zlib --libexecdir=/usr/lib
--without-included-gettext --enable-threads=posix --enable-nls
--with-gxx-include-dir=/usr/include/c++/4.1.3 --program-suffix=-4.1
--enable-__cxa_atexit --enable-clocale=gnu --enable-libstdcxx-debug
--enable-mpfr --enable-checking=release x86_64-linux-gnu
Thread model: posix
gcc version 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)

yutec at yutec-laptop:~$ g77 --version
GNU Fortran (GCC) 3.4.6 (Ubuntu 3.4.6-6ubuntu2)
Copyright (C) 2006 Free Software Foundation, Inc.

GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
You may redistribute copies of GNU Fortran
under the terms of the GNU General Public License.
For more information about these matters, see the file named COPYING
or type the command `info -f g77 Copying'.

yutec at yutec-laptop:~$ python -c 'import sys;print sys.version'
2.5.1 (r251:54863, Oct  5 2007, 13:50:07)
[GCC 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)]

yutec at yutec-laptop:~$ python -c 'import numpy;print numpy.__version__'
1.0.4

yutec at yutec-laptop:~$ cd /usr/lib/python2.5/site-packages/scipy/

yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy$ cd linalg
yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
python setup_atlas_version.py build_ext --inplace --force
Traceback (most recent call last):
  File "setup_atlas_version.py", line 7, in <module>
    from numpy.distutils.misc_util import get_path, default_config_dict
ImportError: cannot import name get_path

yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
python -c 'import atlas_version'
ATLAS version 3.6.0 built by root on Thu Jun  3 21:13:07 UTC 2004:
   UNAME    : Linux ravel 2.6.5 #3 SMP Tue Apr 13 13:41:54 UTC 2004
x86_64 GNU/Linux
   INSTFLG  :
   MMDEF    : /home/camm/atlas3-3.6.0/CONFIG/ARCHS/HAMMER64SSE2/gcc/gemm
   ARCHDEF  : /home/camm/atlas3-3.6.0/CONFIG/ARCHS/HAMMER64SSE2/gcc/misc
   F2CDEFS  : -DAdd__ -DStringSunStyle
   CACHEEDGE: 720896
   F77      : /home/camm/usr/bin/g77, version GNU Fortran (GCC) 3.3.3
(Debian 20040422)
   F77FLAGS : -fomit-frame-pointer -O -m64
   CC       : /usr/bin/gcc, version gcc (GCC) 3.3.3 (Debian 20040422)
   CC FLAGS : -fomit-frame-pointer -O -mfpmath=387 -m64
   MCC      : /usr/bin/gcc, version gcc (GCC) 3.3.3 (Debian 20040422)
   MCCFLAGS : -fomit-frame-pointer -O -mfpmath=387 -m64

yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$ cd ..
yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy$ cd ..
yutec at yutec-laptop:/usr/lib/python2.5/site-packages$ cd numpy
yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy$ cd distutils/

yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy/distutils$
python system_info.py
lapack_info:
  libraries lapack not found in /usr/local/lib
  FOUND:
    libraries = ['lapack']
    library_dirs = ['/usr/lib']
    language = f77

lapack_opt_info:
lapack_mkl_info:
mkl_info:
  libraries mkl,vml,guide not found in /usr/local/lib
  libraries mkl,vml,guide not found in /usr/lib
  NOT AVAILABLE

  NOT AVAILABLE

atlas_threads_info:
Setting PTATLAS=ATLAS
  libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib
  libraries lapack_atlas not found in /usr/local/lib
  libraries ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
  libraries lapack_atlas not found in /usr/lib/atlas
  libraries ptf77blas,ptcblas,atlas not found in /usr/lib
__main__.atlas_threads_info
  NOT AVAILABLE

atlas_info:
  libraries f77blas,cblas,atlas not found in /usr/local/lib
  libraries lapack_atlas not found in /usr/local/lib
  libraries f77blas,cblas,atlas not found in /usr/lib/atlas
  libraries lapack_atlas not found in /usr/lib/atlas
__main__.atlas_info
  FOUND:
    libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
    library_dirs = ['/usr/lib']
    language = f77
    include_dirs = ['/usr/include']

customize GnuFCompiler
Found executable /usr/bin/g77
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler using config
  FOUND:
    libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
    library_dirs = ['/usr/lib']
    language = f77
    define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
    include_dirs = ['/usr/include']

lapack_atlas_info:
  libraries lapack_atlas,f77blas,cblas,atlas not found in /usr/local/lib
  libraries lapack_atlas not found in /usr/local/lib
  libraries lapack_atlas,f77blas,cblas,atlas not found in /usr/lib/atlas
  libraries lapack_atlas not found in /usr/lib/atlas
__main__.lapack_atlas_info
  FOUND:
    libraries = ['lapack', 'lapack_atlas', 'f77blas', 'cblas', 'atlas']
    library_dirs = ['/usr/lib']
    language = f77
    include_dirs = ['/usr/include']

umfpack_info:
  libraries umfpack not found in /usr/local/lib
  libraries umfpack not found in /usr/lib
  NOT AVAILABLE

_pkg_config_info:
Found executable /usr/bin/pkg-config
  NOT AVAILABLE

lapack_atlas_threads_info:
Setting PTATLAS=ATLAS
  libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/local/lib
  libraries lapack_atlas not found in /usr/local/lib
  libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
  libraries lapack_atlas not found in /usr/lib/atlas
  libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/lib
__main__.lapack_atlas_threads_info
  NOT AVAILABLE

x11_info:
  libraries X11 not found in /usr/lib64
  NOT AVAILABLE

blas_info:
  libraries blas not found in /usr/local/lib
  FOUND:
    libraries = ['blas']
    library_dirs = ['/usr/lib']
    language = f77

fftw_info:
  libraries fftw3 not found in /usr/local/lib
  FOUND:
    libraries = ['fftw3']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_FFTW3_H', None)]
    include_dirs = ['/usr/include']

f2py_info:
  FOUND:
    sources = ['/usr/lib/python2.5/site-packages/numpy/f2py/src/fortranobject.c']
    include_dirs = ['/usr/lib/python2.5/site-packages/numpy/f2py/src']

gdk_pixbuf_xlib_2_info:
  NOT AVAILABLE

dfftw_threads_info:
  libraries drfftw_threads,dfftw_threads not found in /usr/local/lib
  libraries drfftw_threads,dfftw_threads not found in /usr/lib
  dfftw threads not found
  NOT AVAILABLE

atlas_blas_info:
  libraries f77blas,cblas,atlas not found in /usr/local/lib
  libraries f77blas,cblas,atlas not found in /usr/lib/atlas
  FOUND:
    libraries = ['f77blas', 'cblas', 'atlas']
    library_dirs = ['/usr/lib']
    language = c
    include_dirs = ['/usr/include']

fftw3_info:
  libraries fftw3 not found in /usr/local/lib
  FOUND:
    libraries = ['fftw3']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_FFTW3_H', None)]
    include_dirs = ['/usr/include']

blas_opt_info:
blas_mkl_info:
  libraries mkl,vml,guide not found in /usr/local/lib
  libraries mkl,vml,guide not found in /usr/lib
  NOT AVAILABLE

atlas_blas_threads_info:
Setting PTATLAS=ATLAS
  libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib
  libraries ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
  libraries ptf77blas,ptcblas,atlas not found in /usr/lib
  NOT AVAILABLE

customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler using config
  FOUND:
    libraries = ['f77blas', 'cblas', 'atlas']
    library_dirs = ['/usr/lib']
    language = c
    define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
    include_dirs = ['/usr/include']

sfftw_info:
  libraries srfftw,sfftw not found in /usr/local/lib
  libraries srfftw,sfftw not found in /usr/lib
  sfftw not found
  NOT AVAILABLE

xft_info:
  NOT AVAILABLE

fft_opt_info:
djbfft_info:
  NOT AVAILABLE

  FOUND:
    libraries = ['fftw3']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_FFTW3_H', None)]
    include_dirs = ['/usr/include']

gdk_x11_2_info:
  NOT AVAILABLE

agg2_info:
  NOT AVAILABLE

numarray_info:
  NOT AVAILABLE

blas_src_info:
  NOT AVAILABLE

fftw2_info:
  libraries rfftw,fftw not found in /usr/local/lib
  FOUND:
    libraries = ['rfftw', 'fftw']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_FFTW_H', None)]
    include_dirs = ['/usr/include']

fftw_threads_info:
  libraries rfftw_threads,fftw_threads not found in /usr/local/lib
  FOUND:
    libraries = ['rfftw_threads', 'fftw_threads']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_FFTW_THREADS_H', None)]
    include_dirs = ['/usr/include']

_numpy_info:
  FOUND:
    define_macros = [('NUMERIC_VERSION', '"\\"24.2\\""'), ('NUMERIC', None)]
    include_dirs = ['/usr/include/python2.5']

wx_info:
Could not locate executable wx-config
File not found: None. Cannot determine wx info.
  NOT AVAILABLE

gdk_info:
  NOT AVAILABLE

gtkp_x11_2_info:
  NOT AVAILABLE

sfftw_threads_info:
  libraries srfftw_threads,sfftw_threads not found in /usr/local/lib
  libraries srfftw_threads,sfftw_threads not found in /usr/lib
  sfftw threads not found
  NOT AVAILABLE

boost_python_info:
  NOT AVAILABLE

freetype2_info:
  NOT AVAILABLE

gdk_2_info:
  NOT AVAILABLE

dfftw_info:
  libraries drfftw,dfftw not found in /usr/local/lib
  libraries drfftw,dfftw not found in /usr/lib
  dfftw not found
  NOT AVAILABLE

lapack_src_info:
  NOT AVAILABLE

gtkp_2_info:
  NOT AVAILABLE

gdk_pixbuf_2_info:
  NOT AVAILABLE

amd_info:
  libraries amd not found in /usr/local/lib
  libraries amd not found in /usr/lib
  NOT AVAILABLE

Numeric_info:
  FOUND:
    define_macros = [('NUMERIC_VERSION', '"\\"24.2\\""'), ('NUMERIC', None)]
    include_dirs = ['/usr/include/python2.5']

numerix_info:
numpy_info:
  FOUND:
    define_macros = [('NUMPY_VERSION', '"\\"1.0.4\\""'), ('NUMPY', None)]

  FOUND:
    define_macros = [('NUMPY_VERSION', '"\\"1.0.4\\""'), ('NUMPY', None)]


From Laurence.Viry at imag.fr  Fri Nov 23 13:16:50 2007
From: Laurence.Viry at imag.fr (Laurence Viry)
Date: Fri, 23 Nov 2007 19:16:50 +0100
Subject: [SciPy-user] install numpy/scipy with Intel Compilers
Message-ID: <47471912.3050300@imag.fr>

Hi!

I try to install numpy 1.0.4 and scipy 0.6.0 with INTEL C/C++ and 
Fortran compilers and SCSL scientific library with python2.4.4 on 
Itanium with linux OS. (Altix SGI with SUSE linux)
 
I didn't success to configure the construct of build_ext for scipy with 
the fortran INTEL compiler (--fcompiler=intele)

* First, I set LD_LIBRARY_PATH and LD_RUN_PATH with directories which 
contains c/c++, fortran and SCSL libraries

export 
LD_LIBRARY_PATH=/opt/intel/intel_cc/9.0.024/lib:/opt/intel/intel_fc/9.0.021/lib:/usr/lib

Export 
LD_RUN_PATH=/opt/intel/intel_cc/9.0.024/lib:/opt/intel/intel_fc/9.0.021/lib:/usr/lib

* I put in the file "site.cfg" :

[DEFAULT]
library_dirs = /usr/lib:/usr/local/lib/:/usr/local/stow/fftw-3.1.2/lib
include_dirs = 
/usr/include:/usr/local/include:/usr/local/stow/fftw-3.1.2/include

[blas]
library_dirs = /usr/lib
include_dirs=/usr/include
blas_libs= scs

[lapack]
library_dirs = /usr/lib
include_dirs=/usr/include
lapack_libs = scs

[fftw]
libraries = fftw3


* Compile and install numpy with:

python setup.py config --compiler=intel build_clib --compiler=intel 
build_ext --compiler=intel install

It's Ok for numpy, it doesn'tt need fortran compiler

but for scipy

python setup.py -v config_cc --compiler=intele config_fc 
--fcompiler=intele  build_clib  --compiler=intele --fcompiler=intele 
build_ext  --compiler=intele --fcompiler=intele install

It's Ok for build_clib, the python script find c and fortran compiler

 it doesn't find fortran compiler for build_ext, first I had a warning

warning: build_ext: f77_compiler=intele is not available, it was 
available for the construct of build_clib? ????

then

warning: build_ext: extension 'scipy.fftpack._fftpack' has Fortran 
libraries but no Fortran linker found, using default linker

for the construct of build_ext, the python script use the default 
fortran compiler g77 and the construct stoped with the message error:

"extension 'scipy.interpolate.dfitpack' has Fortran sources but no 
Fortran compiler found"

why --fcompiler=intele option of build_ext doesn't operate?????????

I show this option of build_ext command with the python command

python setup.py build_ext --help-fcompiler


I set the variables F77, F90 , F77FLAGS with the intel compiler, I had 
the same error

How can I configure the construct of build_ext for scipy with the Intel 
fortran compiler

The solutions I found on the scipy archives didn't work


Thank for your help



-- 
Laurence Viry

CRIP UJF - Projet MIRAGE ( http://mirage.imag.fr )
Laboratoire de Mod?lisation et Calcul - IMAG

tel: 04 76 63 56 34
fax: 04 76 63 12 63
e-mail: Laurence.Viry at imag.fr



From cohen at slac.stanford.edu  Fri Nov 23 14:30:41 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Fri, 23 Nov 2007 11:30:41 -0800
Subject: [SciPy-user] scipy 0.6 installation problem
In-Reply-To: <11001d3e0711230918q1ab5e650s53c98fe8613420dd@mail.gmail.com>
References: <11001d3e0711230918q1ab5e650s53c98fe8613420dd@mail.gmail.com>
Message-ID: <47472A61.6070607@slac.stanford.edu>

well, I am certainly no expert on scipy, but the results of the test 
don't look that bad! When something really bad has occurred at build 
time, you usually don't pass almost all the tests!

Ran 1728 tests in 3.660s
FAILED (failures=2) <unittest._TextTestResult run=1728 errors=0 failures=2>

As for the one that you don't pass, it seems that you end up with odr evaluations different than the ones the test expects... I think I have the same issue, but it does not point, IMHO, to a failure to build correctly.

Experts will tell you about it better than I can, but from your post it 
seems that you are mostly ok to start enjoying scipy
best,
Johann


Yutec Sun wrote:
> Hello,
> I'd like to ask your favor for helping me install scipy 0.6.
> I have a problem importing scipy. In fact, importing doesn't show an
> error message, but scipy.test produces failures.
> My OS is Ubuntu 7.10 (Gutsy Gibbon) amd64. Following the installation
> instruction, I attach the messages for further information.
> Please understand that I'm new to linux system and scipy, and let me
> know how to successfully install scipy step by step.
> I thank you for your help.
>
> Best regards,
>
> Yutec
>
>
> --------------------------------------------------------------------------------------------------------------
> 1. Error message
> --------------------------------------------------------------------------------------------------------------
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy/distutils$ python
> Python 2.5.1 (r251:54863, Oct  5 2007, 13:50:07)
> [GCC 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>   
>>>> import scipy
>>>> scipy.test(level=1)
>>>>         
>   Found 9/9 tests for scipy.cluster.vq
>   Found 18/18 tests for scipy.fftpack.basic
>   Found 4/4 tests for scipy.fftpack.helper
>   Found 20/20 tests for scipy.fftpack.pseudo_diffs
>   Found 1/1 tests for scipy.integrate
>   Found 10/10 tests for scipy.integrate.quadpack
>   Found 3/3 tests for scipy.integrate.quadrature
>   Found 6/6 tests for scipy.interpolate
>   Found 6/6 tests for scipy.interpolate.fitpack
>   Found 4/4 tests for scipy.io.array_import
>   Found 28/28 tests for scipy.io.mio
>   Found 13/13 tests for scipy.io.mmio
>   Found 5/5 tests for scipy.io.npfile
>   Found 4/4 tests for scipy.io.recaster
>   Found 16/16 tests for scipy.lib.blas
>   Found 128/128 tests for scipy.lib.blas.fblas
>   Found 42/42 tests for scipy.lib.lapack
>   Found 41/41 tests for scipy.linalg.basic
> <module 'scipy.linalg.fblas' from
> '/usr/lib/python2.5/site-packages/scipy/linalg/fblas.so'>
>   Found 16/16 tests for scipy.linalg.blas
>   Found 72/72 tests for scipy.linalg.decomp
>   Found 128/128 tests for scipy.linalg.fblas
>   Found 6/6 tests for scipy.linalg.iterative
>   Found 4/4 tests for scipy.linalg.lapack
>   Found 7/7 tests for scipy.linalg.matfuncs
>   Found 9/9 tests for scipy.linsolve.umfpack
>   Found 2/2 tests for scipy.maxentropy
>   Found 3/3 tests for scipy.misc.pilutil
>   Found 399/399 tests for scipy.ndimage
>   Found 5/5 tests for scipy.odr
>   Found 8/8 tests for scipy.optimize
>   Found 1/1 tests for scipy.optimize.cobyla
>   Found 10/10 tests for scipy.optimize.nonlin
>   Found 4/4 tests for scipy.optimize.zeros
>   Found 5/5 tests for scipy.signal.signaltools
>   Found 4/4 tests for scipy.signal.wavelets
>   Found 152/152 tests for scipy.sparse
>   Found 342/342 tests for scipy.special.basic
>   Found 3/3 tests for scipy.special.spfun_stats
>   Found 107/107 tests for scipy.stats
>   Found 73/73 tests for scipy.stats.distributions
>   Found 10/10 tests for scipy.stats.morestats
>   Found 0/0 tests for __main__
> /usr/lib/python2.5/site-packages/scipy/cluster/vq.py:477: UserWarning:
> One of the clusters is empty. Re-run kmean with a different
> initialization.
>   warnings.warn("One of the clusters is empty. "
> ...exception raised as expected: One of the clusters is empty. Re-run
> kmean with a different initialization.
> ................................................Residual: 1.05006987327e-07
> ..................../usr/lib/python2.5/site-packages/scipy/interpolate/fitpack2.py:458:
> UserWarning:
> The coefficients of the spline returned have been computed as the
> minimal norm least-squares solution of a (numerically) rank deficient
> system (deficiency=7). If deficiency is large, the results may be
> inaccurate. Deficiency may strongly depend on the value of eps.
>   warnings.warn(message)
> ......
> Don't worry about a warning regarding the number of bytes read.
> Warning: 1000000 bytes requested, 20 bytes read.
> .........................................................................caxpy:n=4
> ..caxpy:n=3
> ....ccopy:n=4
> ..ccopy:n=3
> .............cscal:n=4
> ....cswap:n=4
> ..cswap:n=3
> .....daxpy:n=4
> ..daxpy:n=3
> ....dcopy:n=4
> ..dcopy:n=3
> .............dscal:n=4
> ....dswap:n=4
> ..dswap:n=3
> .....saxpy:n=4
> ..saxpy:n=3
> ....scopy:n=4
> ..scopy:n=3
> .............sscal:n=4
> ....sswap:n=4
> ..sswap:n=3
> .....zaxpy:n=4
> ..zaxpy:n=3
> ....zcopy:n=4
> ..zcopy:n=3
> .............zscal:n=4
> ....zswap:n=4
> ..zswap:n=3
> ................................................................................................................................................................................caxpy:n=4
> ..caxpy:n=3
> ....ccopy:n=4
> ..ccopy:n=3
> .............cscal:n=4
> ....cswap:n=4
> ..cswap:n=3
> .....daxpy:n=4
> ..daxpy:n=3
> ....dcopy:n=4
> ..dcopy:n=3
> .............dscal:n=4
> ....dswap:n=4
> ..dswap:n=3
> .....saxpy:n=4
> ..saxpy:n=3
> ....scopy:n=4
> ..scopy:n=3
> .............sscal:n=4
> ....sswap:n=4
> ..sswap:n=3
> .....zaxpy:n=4
> ..zaxpy:n=3
> ....zcopy:n=4
> ..zcopy:n=3
> .............zscal:n=4
> ....zswap:n=4
> ..zswap:n=3
> .............Result may be inaccurate, approximate err = 1.56163357201e-08
> ...Result may be inaccurate, approximate err = 1.25221600096e-10
> ......Use minimum degree ordering on A'+A.
> ..Use minimum degree ordering on A'+A.
> ...Use minimum degree ordering on A'+A.
> ............................................................................................................./usr/lib/python2.5/site-packages/scipy/ndimage/interpolation.py:41:
> UserWarning: Mode "reflect" may yield incorrect results on boundaries.
> Please use "mirror" instead.
>   warnings.warn('Mode "reflect" may yield incorrect results on '
> ........................................................................................................................................................................................................................................................................................................F..F.........................................................Use
> minimum degree ordering on A'+A.
> .....................................Use minimum degree ordering on A'+A.
> .....................................Use minimum degree ordering on A'+A.
> ................................Use minimum degree ordering on A'+A.
> ....................................................................................................................................................................................................................................................................................................................................................0.2
> 0.2
> 0.2
> ......0.2
> ..0.2
> 0.2
> 0.2
> 0.2
> 0.2
> .........................................................................................................................................................................................................Ties
> preclude use of exact statistic.
> ..Ties preclude use of exact statistic.
> ......
> ======================================================================
> FAIL: test_explicit (scipy.tests.test_odr.test_odr)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>   File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
> line 50, in test_explicit
>     -8.7849712165253724e-02]),
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 232, in assert_array_almost_equal
>     header='Arrays are not almost equal')
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 217, in assert_array_compare
>     assert cond, msg
> AssertionError:
> Arrays are not almost equal
>
> (mismatch 100.0%)
>  x: array([  1.26462971e+03,  -5.42545890e+01,  -8.64250389e-02])
>  y: array([  1.26465481e+03,  -5.40184100e+01,  -8.78497122e-02])
>
> ======================================================================
> FAIL: test_multi (scipy.tests.test_odr.test_odr)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>   File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
> line 191, in test_multi
>     0.5101147161764654,  0.5173902330489161]),
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 232, in assert_array_almost_equal
>     header='Arrays are not almost equal')
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 217, in assert_array_compare
>     assert cond, msg
> AssertionError:
> Arrays are not almost equal
>
> (mismatch 100.0%)
>  x: array([ 4.31272063,  2.44289312,  7.76215871,  0.55995622,  0.46423343])
>  y: array([ 4.37998803,  2.43330576,  8.00288459,  0.51011472,  0.51739023])
>
> ----------------------------------------------------------------------
> Ran 1728 tests in 3.660s
>
> FAILED (failures=2)
> <unittest._TextTestResult run=1728 errors=0 failures=2>
>
>
>
> --------------------------------------------------------------------------------------------------------------
> 2. System check
> --------------------------------------------------------------------------------------------------------------
> yutec at yutec-laptop:~$ python -c 'import os,sys;print os.name,sys.platform'
> posix linux2
>
> yutec at yutec-laptop:~$ uname -a
> Linux yutec-laptop 2.6.22-14-generic #1 SMP Sun Oct 14 21:45:15 GMT
> 2007 x86_64 GNU/Linux
>
> yutec at yutec-laptop:~$ gcc -v
> Using built-in specs.
> Target: x86_64-linux-gnu
> Configured with: ../src/configure -v
> --enable-languages=c,c++,fortran,objc,obj-c++,treelang --prefix=/usr
> --enable-shared --with-system-zlib --libexecdir=/usr/lib
> --without-included-gettext --enable-threads=posix --enable-nls
> --with-gxx-include-dir=/usr/include/c++/4.1.3 --program-suffix=-4.1
> --enable-__cxa_atexit --enable-clocale=gnu --enable-libstdcxx-debug
> --enable-mpfr --enable-checking=release x86_64-linux-gnu
> Thread model: posix
> gcc version 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)
>
> yutec at yutec-laptop:~$ g77 --version
> GNU Fortran (GCC) 3.4.6 (Ubuntu 3.4.6-6ubuntu2)
> Copyright (C) 2006 Free Software Foundation, Inc.
>
> GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
> You may redistribute copies of GNU Fortran
> under the terms of the GNU General Public License.
> For more information about these matters, see the file named COPYING
> or type the command `info -f g77 Copying'.
>
> yutec at yutec-laptop:~$ python -c 'import sys;print sys.version'
> 2.5.1 (r251:54863, Oct  5 2007, 13:50:07)
> [GCC 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)]
>
> yutec at yutec-laptop:~$ python -c 'import numpy;print numpy.__version__'
> 1.0.4
>
> yutec at yutec-laptop:~$ cd /usr/lib/python2.5/site-packages/scipy/
>
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy$ cd linalg
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
> python setup_atlas_version.py build_ext --inplace --force
> Traceback (most recent call last):
>   File "setup_atlas_version.py", line 7, in <module>
>     from numpy.distutils.misc_util import get_path, default_config_dict
> ImportError: cannot import name get_path
>
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
> python -c 'import atlas_version'
> ATLAS version 3.6.0 built by root on Thu Jun  3 21:13:07 UTC 2004:
>    UNAME    : Linux ravel 2.6.5 #3 SMP Tue Apr 13 13:41:54 UTC 2004
> x86_64 GNU/Linux
>    INSTFLG  :
>    MMDEF    : /home/camm/atlas3-3.6.0/CONFIG/ARCHS/HAMMER64SSE2/gcc/gemm
>    ARCHDEF  : /home/camm/atlas3-3.6.0/CONFIG/ARCHS/HAMMER64SSE2/gcc/misc
>    F2CDEFS  : -DAdd__ -DStringSunStyle
>    CACHEEDGE: 720896
>    F77      : /home/camm/usr/bin/g77, version GNU Fortran (GCC) 3.3.3
> (Debian 20040422)
>    F77FLAGS : -fomit-frame-pointer -O -m64
>    CC       : /usr/bin/gcc, version gcc (GCC) 3.3.3 (Debian 20040422)
>    CC FLAGS : -fomit-frame-pointer -O -mfpmath=387 -m64
>    MCC      : /usr/bin/gcc, version gcc (GCC) 3.3.3 (Debian 20040422)
>    MCCFLAGS : -fomit-frame-pointer -O -mfpmath=387 -m64
>
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$ cd ..
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy$ cd ..
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages$ cd numpy
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy$ cd distutils/
>
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy/distutils$
> python system_info.py
> lapack_info:
>   libraries lapack not found in /usr/local/lib
>   FOUND:
>     libraries = ['lapack']
>     library_dirs = ['/usr/lib']
>     language = f77
>
> lapack_opt_info:
> lapack_mkl_info:
> mkl_info:
>   libraries mkl,vml,guide not found in /usr/local/lib
>   libraries mkl,vml,guide not found in /usr/lib
>   NOT AVAILABLE
>
>   NOT AVAILABLE
>
> atlas_threads_info:
> Setting PTATLAS=ATLAS
>   libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib
> __main__.atlas_threads_info
>   NOT AVAILABLE
>
> atlas_info:
>   libraries f77blas,cblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries f77blas,cblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
> __main__.atlas_info
>   FOUND:
>     libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = f77
>     include_dirs = ['/usr/include']
>
> customize GnuFCompiler
> Found executable /usr/bin/g77
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler using config
>   FOUND:
>     libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = f77
>     define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
>     include_dirs = ['/usr/include']
>
> lapack_atlas_info:
>   libraries lapack_atlas,f77blas,cblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries lapack_atlas,f77blas,cblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
> __main__.lapack_atlas_info
>   FOUND:
>     libraries = ['lapack', 'lapack_atlas', 'f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = f77
>     include_dirs = ['/usr/include']
>
> umfpack_info:
>   libraries umfpack not found in /usr/local/lib
>   libraries umfpack not found in /usr/lib
>   NOT AVAILABLE
>
> _pkg_config_info:
> Found executable /usr/bin/pkg-config
>   NOT AVAILABLE
>
> lapack_atlas_threads_info:
> Setting PTATLAS=ATLAS
>   libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
>   libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/lib
> __main__.lapack_atlas_threads_info
>   NOT AVAILABLE
>
> x11_info:
>   libraries X11 not found in /usr/lib64
>   NOT AVAILABLE
>
> blas_info:
>   libraries blas not found in /usr/local/lib
>   FOUND:
>     libraries = ['blas']
>     library_dirs = ['/usr/lib']
>     language = f77
>
> fftw_info:
>   libraries fftw3 not found in /usr/local/lib
>   FOUND:
>     libraries = ['fftw3']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW3_H', None)]
>     include_dirs = ['/usr/include']
>
> f2py_info:
>   FOUND:
>     sources = ['/usr/lib/python2.5/site-packages/numpy/f2py/src/fortranobject.c']
>     include_dirs = ['/usr/lib/python2.5/site-packages/numpy/f2py/src']
>
> gdk_pixbuf_xlib_2_info:
>   NOT AVAILABLE
>
> dfftw_threads_info:
>   libraries drfftw_threads,dfftw_threads not found in /usr/local/lib
>   libraries drfftw_threads,dfftw_threads not found in /usr/lib
>   dfftw threads not found
>   NOT AVAILABLE
>
> atlas_blas_info:
>   libraries f77blas,cblas,atlas not found in /usr/local/lib
>   libraries f77blas,cblas,atlas not found in /usr/lib/atlas
>   FOUND:
>     libraries = ['f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = c
>     include_dirs = ['/usr/include']
>
> fftw3_info:
>   libraries fftw3 not found in /usr/local/lib
>   FOUND:
>     libraries = ['fftw3']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW3_H', None)]
>     include_dirs = ['/usr/include']
>
> blas_opt_info:
> blas_mkl_info:
>   libraries mkl,vml,guide not found in /usr/local/lib
>   libraries mkl,vml,guide not found in /usr/lib
>   NOT AVAILABLE
>
> atlas_blas_threads_info:
> Setting PTATLAS=ATLAS
>   libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib
>   NOT AVAILABLE
>
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler using config
>   FOUND:
>     libraries = ['f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = c
>     define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
>     include_dirs = ['/usr/include']
>
> sfftw_info:
>   libraries srfftw,sfftw not found in /usr/local/lib
>   libraries srfftw,sfftw not found in /usr/lib
>   sfftw not found
>   NOT AVAILABLE
>
> xft_info:
>   NOT AVAILABLE
>
> fft_opt_info:
> djbfft_info:
>   NOT AVAILABLE
>
>   FOUND:
>     libraries = ['fftw3']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW3_H', None)]
>     include_dirs = ['/usr/include']
>
> gdk_x11_2_info:
>   NOT AVAILABLE
>
> agg2_info:
>   NOT AVAILABLE
>
> numarray_info:
>   NOT AVAILABLE
>
> blas_src_info:
>   NOT AVAILABLE
>
> fftw2_info:
>   libraries rfftw,fftw not found in /usr/local/lib
>   FOUND:
>     libraries = ['rfftw', 'fftw']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW_H', None)]
>     include_dirs = ['/usr/include']
>
> fftw_threads_info:
>   libraries rfftw_threads,fftw_threads not found in /usr/local/lib
>   FOUND:
>     libraries = ['rfftw_threads', 'fftw_threads']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW_THREADS_H', None)]
>     include_dirs = ['/usr/include']
>
> _numpy_info:
>   FOUND:
>     define_macros = [('NUMERIC_VERSION', '"\\"24.2\\""'), ('NUMERIC', None)]
>     include_dirs = ['/usr/include/python2.5']
>
> wx_info:
> Could not locate executable wx-config
> File not found: None. Cannot determine wx info.
>   NOT AVAILABLE
>
> gdk_info:
>   NOT AVAILABLE
>
> gtkp_x11_2_info:
>   NOT AVAILABLE
>
> sfftw_threads_info:
>   libraries srfftw_threads,sfftw_threads not found in /usr/local/lib
>   libraries srfftw_threads,sfftw_threads not found in /usr/lib
>   sfftw threads not found
>   NOT AVAILABLE
>
> boost_python_info:
>   NOT AVAILABLE
>
> freetype2_info:
>   NOT AVAILABLE
>
> gdk_2_info:
>   NOT AVAILABLE
>
> dfftw_info:
>   libraries drfftw,dfftw not found in /usr/local/lib
>   libraries drfftw,dfftw not found in /usr/lib
>   dfftw not found
>   NOT AVAILABLE
>
> lapack_src_info:
>   NOT AVAILABLE
>
> gtkp_2_info:
>   NOT AVAILABLE
>
> gdk_pixbuf_2_info:
>   NOT AVAILABLE
>
> amd_info:
>   libraries amd not found in /usr/local/lib
>   libraries amd not found in /usr/lib
>   NOT AVAILABLE
>
> Numeric_info:
>   FOUND:
>     define_macros = [('NUMERIC_VERSION', '"\\"24.2\\""'), ('NUMERIC', None)]
>     include_dirs = ['/usr/include/python2.5']
>
> numerix_info:
> numpy_info:
>   FOUND:
>     define_macros = [('NUMPY_VERSION', '"\\"1.0.4\\""'), ('NUMPY', None)]
>
>   FOUND:
>     define_macros = [('NUMPY_VERSION', '"\\"1.0.4\\""'), ('NUMPY', None)]
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From s.mientki at ru.nl  Fri Nov 23 15:16:48 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Fri, 23 Nov 2007 21:16:48 +0100
Subject: [SciPy-user] this must sound stupid ... (how to delete and destroy
 an AUI-pane ?)
Message-ID: <47473530.7060200@ru.nl>

hello,

I want to dynamically create, modify and delete AUI-panes in a 
AUI_DockingWindowMgr,

Now I can create panes:
    self._mgr.AddPane( self.CreateHTMLCtrl(), wx.aui.AuiPaneInfo().
                       Name("test3").Caption("Client Size Reporter").
                       Left().
                       CloseButton ( False ).
                       MaximizeButton ( False ) )

I can find the pane again
    pane = self._mgr.GetPane("test2")

but how do I delete and destroy  a pane ?
The next statements all returns errors :-(
    pane.Close()
    self._mgr.DetachPane(pane)

thanks,
Stef


From cohen at slac.stanford.edu  Fri Nov 23 15:11:38 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Fri, 23 Nov 2007 12:11:38 -0800
Subject: [SciPy-user] Building with a non-standard gfortran
In-Reply-To: <4BD1DAD9-E9B8-4AE2-93E5-FE46B146830D@physics.mcmaster.ca>
References: <799406d60711200603t8e451cah2e13add2bd794a11@mail.gmail.com>
	<4BD1DAD9-E9B8-4AE2-93E5-FE46B146830D@physics.mcmaster.ca>
Message-ID: <474733FA.1000108@slac.stanford.edu>

David M. Cooke wrote:
> On Nov 20, 2007, at 09:03 , Adam Mercer wrote:
>
>   
>> Hi
>>
>> Using MacPorts, the gfortran compiler, from gcc-4.2.x, is installed as
>> gfortran-mp-4.2.  I have been trying to build scipy with this compiler
>> and found that that numpy I need to build and install with
>>
>> $ python setup.py config_fc --fcompiler gnu95 \
>>    --f77exec /opt/local/bin/gfortran-mp-4.2 \
>>    --f90exec /opt/local/bin/gfortran-mp-4.2 build
>> $ python setup.py install --prefix=${NUMPY_LOCATION}
>>
>> however using the above for scipy fails on the install phase with it
>> being unable to locate a fortran compiler.  I therefore need to
>> install scipy with
>>
>> $ python setup.py config_fc --fcompiler gnu95 \
>>    --f77exec /opt/local/bin/gfortran-mp-4.2 \
>>    --f90exec /opt/local/bin/gfortran-mp-4.2 install --prefix=$ 
>> {SCIPY_LOCATION}
>>
>> Is this expected, as I would have expected install to inherit the
>> options that where used during the build phase?
>>     
>
> The options used in the build phase aren't stored, so, yes, they need  
> to be respecified. For numpy it doesn't matter, as the Fortran  
> compiler isn't used. Since we use distutils, you can set the config_fc  
> options in a file (either globally for the user in a  
> ~/.pydistutils.cfg, or per-package in a setup.cfg next to the setup.py  
> you're running) by adding the following section to the appropiate file
>
> [config_fc]
> fcompiler=gfortran
> f77exec=gfortran-mp-4.2
> f90exec=gfortran-mp-4.2
>
>   
Thanks for the tip, David. I tried it :
[cohen at localhost scipy-svn]$ which gfortran
/usr/bin/gfortran
[cohen at localhost scipy-svn]$ more ~/.pydistutils.cfg
[config_fc]
fcompiler=gfortran
f77exec=gfortran
f90exec=gfortran

and 'python setup.py build' does correctly use gfortran, but both numpy 
and scipy keep putting out
customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler using build_ext

at install time....

Also, now that I am at it, there seems to be a doubling of the mach 
library, which should be solved for cleanliness :
customize UnixCCompiler using build_ext
library 'mach' defined more than once, overwriting build_info
{'sources': ['scipy/integrate/mach/i1mach.f', 
'scipy/integrate/mach/xerror.f', 'scipy/integrate/mach/r1mach.f', 
'scipy/integrate/mach/d1mach.f'], 'config_fc': {'noopt': 
('scipy/integrate/setup.pyc', 1)}, 'source_languages': ['f77']}...
with
{'sources': ['scipy/special/mach/i1mach.f', 
'scipy/special/mach/xerror.f', 'scipy/special/mach/r1mach.f', 
'scipy/special/mach/d1mach.f'], 'config_fc': {'noopt': 
('scipy/special/setup.pyc', 1)}, 'source_languages': ['f77']}...


best,
Johann


From s.mientki at ru.nl  Fri Nov 23 15:20:09 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Fri, 23 Nov 2007 21:20:09 +0100
Subject: [SciPy-user] sorry wrong list,
 was: this must sound stupid ... (how to delete and destroy an
 AUI-pane ?)
In-Reply-To: <47473530.7060200@ru.nl>
References: <47473530.7060200@ru.nl>
Message-ID: <474735F9.2050203@ru.nl>

sorry wrong list ;-)

Stef Mientki wrote:
> hello,
>
> I want to dynamically create, modify and delete AUI-panes in a 
> AUI_DockingWindowMgr,
>
> Now I can create panes:
>     self._mgr.AddPane( self.CreateHTMLCtrl(), wx.aui.AuiPaneInfo().
>                        Name("test3").Caption("Client Size Reporter").
>                        Left().
>                        CloseButton ( False ).
>                        MaximizeButton ( False ) )
>
> I can find the pane again
>     pane = self._mgr.GetPane("test2")
>
> but how do I delete and destroy  a pane ?
> The next statements all returns errors :-(
>     pane.Close()
>     self._mgr.DetachPane(pane)
>
> thanks,
> Stef
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>   


From gnurser at googlemail.com  Fri Nov 23 18:40:34 2007
From: gnurser at googlemail.com (George Nurser)
Date: Fri, 23 Nov 2007 23:40:34 +0000
Subject: [SciPy-user] Building with a non-standard gfortran
In-Reply-To: <4BD1DAD9-E9B8-4AE2-93E5-FE46B146830D@physics.mcmaster.ca>
References: <799406d60711200603t8e451cah2e13add2bd794a11@mail.gmail.com>
	<4BD1DAD9-E9B8-4AE2-93E5-FE46B146830D@physics.mcmaster.ca>
Message-ID: <1d1e6ea70711231540k219f0b17h864ca472a2e46805@mail.gmail.com>

I've got an extended version of the same query.
I have installed gcc43 into /opt/local/bin from macports, and would
like to use this gcc (and associated c++ and gfortran) to compile
numpy/scipy/pytables/matplotlib.

/opt/local/bin is at the start of my $PATH

I understand I create ~/.pydistutils.cfg
something like

[config_fc]
fcompiler=gfortran
f77exec=gfortran-mp-4.3
f90exec=gfortran-mp-4.3
opt='-mtune=core2  -march=core2'

[config]
compiler=gcc
cc=gcc-mp-4.3


But ...
1. how do i specify the gcc4.3 C++ compiler for matplotlib etc..?
2. how do i specify the C & C++ compiler options (like -mtune=core2,
-march=core2)
2. do I need to add /opt/local/lib to my $LD_LIBRARY_PATH? Or is there
a less drastic method?
3. Is there anything else I need?

Apologies if I've missed the documentation.

Regards, George Nurser.


From markbak at gmail.com  Sat Nov 24 08:05:07 2007
From: markbak at gmail.com (Mark Bakker)
Date: Sat, 24 Nov 2007 14:05:07 +0100
Subject: [SciPy-user] Installation of Python/mpl/numpy/scipy/Ipython for
	students
Message-ID: <6946b9500711240505n711f66c6m5e2b63db294a1957@mail.gmail.com>

A question was posted a week or so ago on the scipy list about installing
Python/mpl/numpy/scipy/Ipython by David Arnold (both Windows and Mac). He
was interested in a simple install for students (and others that are less
computer-savy). We had a little off-list correspondence, but thought to
bring it back to the list, also posting on mpl (cause I read that list and
it seems pretty responsive) and the numpy Google discussion group (they
respond VERY quickly).

On the scipy list it was suggested to use Sage reply, but that sounds like a
little overkill to me.

I have used the enthought package on windows successfully in class in
the past, but I understand that they have now switched to eggs, which
is not convenient (yet).

I also don't like all the other things I have to set myself. A switch
on the IDLE GUI to get mpl to work interactively (which is now seems
impossible to do), and setting my PYTHONPATH with environment variables. It
is all easy to do, but not for students, especially students without much
knowledge of how computers work let alone access to set these variables and
options.

At one point I was reading about making a superpack (or a name like
that) installation for Python with our favorite packages as mentioned above.
I am not sure where that is going.

Are others dreaming about such a superpack? What needs to be done to get it
going?

Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071124/472ca173/attachment.html>

From anand.prabhakar.patil at gmail.com  Sat Nov 24 11:39:49 2007
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Sat, 24 Nov 2007 08:39:49 -0800
Subject: [SciPy-user] sparse BLAS in scipy?
Message-ID: <2bc7a5a50711240839m19024b4aw9f73f4217f9a27ec@mail.gmail.com>

Hi all,

Is there a set of python-callable sparse BLAS out there yet? I haven't
found one, and not for lack of Googling. I'm willing to work on
swigging a library if the need exists, but would like some guidance
from someone with more experience with these things.

First, it looks like there are three implementations that would be
good for this purpose:

- NIST sparse BLAS
- SparseLib++ (superseded by above?)
- Boost's ublas with sparse template parameters.

Questions:
- Do I need to do this, or are wrappers already available?
- If so, which library would be best? I lean toward Boost, just
because it's so broadly templatized that scripting a wrapper for all
all the sparse-sparse and sparse-dense versions should be relatively
easy.
- What should the calling conventions from Python be like?
- Any other pointers? (things I should know about numpy.i, for example).

Anand


From lev at vpac.org  Sun Nov 25 17:11:57 2007
From: lev at vpac.org (Lev Lafayette)
Date: Mon, 26 Nov 2007 09:11:57 +1100
Subject: [SciPy-user] scipy 0.6 installation problem
In-Reply-To: <11001d3e0711230918q1ab5e650s53c98fe8613420dd@mail.gmail.com>
References: <11001d3e0711230918q1ab5e650s53c98fe8613420dd@mail.gmail.com>
Message-ID: <1196028717.4934.9.camel@sys09.in.vpac.org>

On Fri, 2007-11-23 at 12:18 -0500, Yutec Sun wrote:
> Hello,
> I'd like to ask your favor for helping me install scipy 0.6.
> I have a problem importing scipy. In fact, importing doesn't show an
> error message, but scipy.test produces failures.
> My OS is Ubuntu 7.10 (Gutsy Gibbon) amd64. Following the installation
> instruction, I attach the messages for further information.
> Please understand that I'm new to linux system and scipy, and let me
> know how to successfully install scipy step by step.
> I thank you for your help.
> 

64bit, eh?

Have you done these modifications?

http://scipy.org/scipy/scipy/changeset/3498

All the best,


Lev



From strawman at astraw.com  Mon Nov 26 00:35:55 2007
From: strawman at astraw.com (Andrew Straw)
Date: Sun, 25 Nov 2007 21:35:55 -0800
Subject: [SciPy-user] scipy 0.6 installation problem
In-Reply-To: <11001d3e0711230918q1ab5e650s53c98fe8613420dd@mail.gmail.com>
References: <11001d3e0711230918q1ab5e650s53c98fe8613420dd@mail.gmail.com>
Message-ID: <474A5B3B.40408@astraw.com>

Hi Yutec,

You can install my version:

http://debs.astraw.com/gutsy/python-scipy_0.6.0-1ubuntu0~0ads2_amd64.deb

-Andrew

Yutec Sun wrote:
> Hello,
> I'd like to ask your favor for helping me install scipy 0.6.
> I have a problem importing scipy. In fact, importing doesn't show an
> error message, but scipy.test produces failures.
> My OS is Ubuntu 7.10 (Gutsy Gibbon) amd64. Following the installation
> instruction, I attach the messages for further information.
> Please understand that I'm new to linux system and scipy, and let me
> know how to successfully install scipy step by step.
> I thank you for your help.
> 
> Best regards,
> 
> Yutec
> 
> 
> --------------------------------------------------------------------------------------------------------------
> 1. Error message
> --------------------------------------------------------------------------------------------------------------
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy/distutils$ python
> Python 2.5.1 (r251:54863, Oct  5 2007, 13:50:07)
> [GCC 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>>>> import scipy
>>>> scipy.test(level=1)
>   Found 9/9 tests for scipy.cluster.vq
>   Found 18/18 tests for scipy.fftpack.basic
>   Found 4/4 tests for scipy.fftpack.helper
>   Found 20/20 tests for scipy.fftpack.pseudo_diffs
>   Found 1/1 tests for scipy.integrate
>   Found 10/10 tests for scipy.integrate.quadpack
>   Found 3/3 tests for scipy.integrate.quadrature
>   Found 6/6 tests for scipy.interpolate
>   Found 6/6 tests for scipy.interpolate.fitpack
>   Found 4/4 tests for scipy.io.array_import
>   Found 28/28 tests for scipy.io.mio
>   Found 13/13 tests for scipy.io.mmio
>   Found 5/5 tests for scipy.io.npfile
>   Found 4/4 tests for scipy.io.recaster
>   Found 16/16 tests for scipy.lib.blas
>   Found 128/128 tests for scipy.lib.blas.fblas
>   Found 42/42 tests for scipy.lib.lapack
>   Found 41/41 tests for scipy.linalg.basic
> <module 'scipy.linalg.fblas' from
> '/usr/lib/python2.5/site-packages/scipy/linalg/fblas.so'>
>   Found 16/16 tests for scipy.linalg.blas
>   Found 72/72 tests for scipy.linalg.decomp
>   Found 128/128 tests for scipy.linalg.fblas
>   Found 6/6 tests for scipy.linalg.iterative
>   Found 4/4 tests for scipy.linalg.lapack
>   Found 7/7 tests for scipy.linalg.matfuncs
>   Found 9/9 tests for scipy.linsolve.umfpack
>   Found 2/2 tests for scipy.maxentropy
>   Found 3/3 tests for scipy.misc.pilutil
>   Found 399/399 tests for scipy.ndimage
>   Found 5/5 tests for scipy.odr
>   Found 8/8 tests for scipy.optimize
>   Found 1/1 tests for scipy.optimize.cobyla
>   Found 10/10 tests for scipy.optimize.nonlin
>   Found 4/4 tests for scipy.optimize.zeros
>   Found 5/5 tests for scipy.signal.signaltools
>   Found 4/4 tests for scipy.signal.wavelets
>   Found 152/152 tests for scipy.sparse
>   Found 342/342 tests for scipy.special.basic
>   Found 3/3 tests for scipy.special.spfun_stats
>   Found 107/107 tests for scipy.stats
>   Found 73/73 tests for scipy.stats.distributions
>   Found 10/10 tests for scipy.stats.morestats
>   Found 0/0 tests for __main__
> /usr/lib/python2.5/site-packages/scipy/cluster/vq.py:477: UserWarning:
> One of the clusters is empty. Re-run kmean with a different
> initialization.
>   warnings.warn("One of the clusters is empty. "
> ...exception raised as expected: One of the clusters is empty. Re-run
> kmean with a different initialization.
> ................................................Residual: 1.05006987327e-07
> ..................../usr/lib/python2.5/site-packages/scipy/interpolate/fitpack2.py:458:
> UserWarning:
> The coefficients of the spline returned have been computed as the
> minimal norm least-squares solution of a (numerically) rank deficient
> system (deficiency=7). If deficiency is large, the results may be
> inaccurate. Deficiency may strongly depend on the value of eps.
>   warnings.warn(message)
> ......
> Don't worry about a warning regarding the number of bytes read.
> Warning: 1000000 bytes requested, 20 bytes read.
> .........................................................................caxpy:n=4
> ..caxpy:n=3
> ....ccopy:n=4
> ..ccopy:n=3
> .............cscal:n=4
> ....cswap:n=4
> ..cswap:n=3
> .....daxpy:n=4
> ..daxpy:n=3
> ....dcopy:n=4
> ..dcopy:n=3
> .............dscal:n=4
> ....dswap:n=4
> ..dswap:n=3
> .....saxpy:n=4
> ..saxpy:n=3
> ....scopy:n=4
> ..scopy:n=3
> .............sscal:n=4
> ....sswap:n=4
> ..sswap:n=3
> .....zaxpy:n=4
> ..zaxpy:n=3
> ....zcopy:n=4
> ..zcopy:n=3
> .............zscal:n=4
> ....zswap:n=4
> ..zswap:n=3
> ................................................................................................................................................................................caxpy:n=4
> ..caxpy:n=3
> ....ccopy:n=4
> ..ccopy:n=3
> .............cscal:n=4
> ....cswap:n=4
> ..cswap:n=3
> .....daxpy:n=4
> ..daxpy:n=3
> ....dcopy:n=4
> ..dcopy:n=3
> .............dscal:n=4
> ....dswap:n=4
> ..dswap:n=3
> .....saxpy:n=4
> ..saxpy:n=3
> ....scopy:n=4
> ..scopy:n=3
> .............sscal:n=4
> ....sswap:n=4
> ..sswap:n=3
> .....zaxpy:n=4
> ..zaxpy:n=3
> ....zcopy:n=4
> ..zcopy:n=3
> .............zscal:n=4
> ....zswap:n=4
> ..zswap:n=3
> .............Result may be inaccurate, approximate err = 1.56163357201e-08
> ...Result may be inaccurate, approximate err = 1.25221600096e-10
> ......Use minimum degree ordering on A'+A.
> ..Use minimum degree ordering on A'+A.
> ...Use minimum degree ordering on A'+A.
> ............................................................................................................./usr/lib/python2.5/site-packages/scipy/ndimage/interpolation.py:41:
> UserWarning: Mode "reflect" may yield incorrect results on boundaries.
> Please use "mirror" instead.
>   warnings.warn('Mode "reflect" may yield incorrect results on '
> ........................................................................................................................................................................................................................................................................................................F..F.........................................................Use
> minimum degree ordering on A'+A.
> .....................................Use minimum degree ordering on A'+A.
> .....................................Use minimum degree ordering on A'+A.
> ................................Use minimum degree ordering on A'+A.
> ....................................................................................................................................................................................................................................................................................................................................................0.2
> 0.2
> 0.2
> ......0.2
> ..0.2
> 0.2
> 0.2
> 0.2
> 0.2
> .........................................................................................................................................................................................................Ties
> preclude use of exact statistic.
> ..Ties preclude use of exact statistic.
> ......
> ======================================================================
> FAIL: test_explicit (scipy.tests.test_odr.test_odr)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>   File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
> line 50, in test_explicit
>     -8.7849712165253724e-02]),
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 232, in assert_array_almost_equal
>     header='Arrays are not almost equal')
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 217, in assert_array_compare
>     assert cond, msg
> AssertionError:
> Arrays are not almost equal
> 
> (mismatch 100.0%)
>  x: array([  1.26462971e+03,  -5.42545890e+01,  -8.64250389e-02])
>  y: array([  1.26465481e+03,  -5.40184100e+01,  -8.78497122e-02])
> 
> ======================================================================
> FAIL: test_multi (scipy.tests.test_odr.test_odr)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>   File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
> line 191, in test_multi
>     0.5101147161764654,  0.5173902330489161]),
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 232, in assert_array_almost_equal
>     header='Arrays are not almost equal')
>   File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
> 217, in assert_array_compare
>     assert cond, msg
> AssertionError:
> Arrays are not almost equal
> 
> (mismatch 100.0%)
>  x: array([ 4.31272063,  2.44289312,  7.76215871,  0.55995622,  0.46423343])
>  y: array([ 4.37998803,  2.43330576,  8.00288459,  0.51011472,  0.51739023])
> 
> ----------------------------------------------------------------------
> Ran 1728 tests in 3.660s
> 
> FAILED (failures=2)
> <unittest._TextTestResult run=1728 errors=0 failures=2>
> 
> 
> 
> --------------------------------------------------------------------------------------------------------------
> 2. System check
> --------------------------------------------------------------------------------------------------------------
> yutec at yutec-laptop:~$ python -c 'import os,sys;print os.name,sys.platform'
> posix linux2
> 
> yutec at yutec-laptop:~$ uname -a
> Linux yutec-laptop 2.6.22-14-generic #1 SMP Sun Oct 14 21:45:15 GMT
> 2007 x86_64 GNU/Linux
> 
> yutec at yutec-laptop:~$ gcc -v
> Using built-in specs.
> Target: x86_64-linux-gnu
> Configured with: ../src/configure -v
> --enable-languages=c,c++,fortran,objc,obj-c++,treelang --prefix=/usr
> --enable-shared --with-system-zlib --libexecdir=/usr/lib
> --without-included-gettext --enable-threads=posix --enable-nls
> --with-gxx-include-dir=/usr/include/c++/4.1.3 --program-suffix=-4.1
> --enable-__cxa_atexit --enable-clocale=gnu --enable-libstdcxx-debug
> --enable-mpfr --enable-checking=release x86_64-linux-gnu
> Thread model: posix
> gcc version 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)
> 
> yutec at yutec-laptop:~$ g77 --version
> GNU Fortran (GCC) 3.4.6 (Ubuntu 3.4.6-6ubuntu2)
> Copyright (C) 2006 Free Software Foundation, Inc.
> 
> GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
> You may redistribute copies of GNU Fortran
> under the terms of the GNU General Public License.
> For more information about these matters, see the file named COPYING
> or type the command `info -f g77 Copying'.
> 
> yutec at yutec-laptop:~$ python -c 'import sys;print sys.version'
> 2.5.1 (r251:54863, Oct  5 2007, 13:50:07)
> [GCC 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)]
> 
> yutec at yutec-laptop:~$ python -c 'import numpy;print numpy.__version__'
> 1.0.4
> 
> yutec at yutec-laptop:~$ cd /usr/lib/python2.5/site-packages/scipy/
> 
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy$ cd linalg
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
> python setup_atlas_version.py build_ext --inplace --force
> Traceback (most recent call last):
>   File "setup_atlas_version.py", line 7, in <module>
>     from numpy.distutils.misc_util import get_path, default_config_dict
> ImportError: cannot import name get_path
> 
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$
> python -c 'import atlas_version'
> ATLAS version 3.6.0 built by root on Thu Jun  3 21:13:07 UTC 2004:
>    UNAME    : Linux ravel 2.6.5 #3 SMP Tue Apr 13 13:41:54 UTC 2004
> x86_64 GNU/Linux
>    INSTFLG  :
>    MMDEF    : /home/camm/atlas3-3.6.0/CONFIG/ARCHS/HAMMER64SSE2/gcc/gemm
>    ARCHDEF  : /home/camm/atlas3-3.6.0/CONFIG/ARCHS/HAMMER64SSE2/gcc/misc
>    F2CDEFS  : -DAdd__ -DStringSunStyle
>    CACHEEDGE: 720896
>    F77      : /home/camm/usr/bin/g77, version GNU Fortran (GCC) 3.3.3
> (Debian 20040422)
>    F77FLAGS : -fomit-frame-pointer -O -m64
>    CC       : /usr/bin/gcc, version gcc (GCC) 3.3.3 (Debian 20040422)
>    CC FLAGS : -fomit-frame-pointer -O -mfpmath=387 -m64
>    MCC      : /usr/bin/gcc, version gcc (GCC) 3.3.3 (Debian 20040422)
>    MCCFLAGS : -fomit-frame-pointer -O -mfpmath=387 -m64
> 
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy/linalg$ cd ..
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/scipy$ cd ..
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages$ cd numpy
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy$ cd distutils/
> 
> yutec at yutec-laptop:/usr/lib/python2.5/site-packages/numpy/distutils$
> python system_info.py
> lapack_info:
>   libraries lapack not found in /usr/local/lib
>   FOUND:
>     libraries = ['lapack']
>     library_dirs = ['/usr/lib']
>     language = f77
> 
> lapack_opt_info:
> lapack_mkl_info:
> mkl_info:
>   libraries mkl,vml,guide not found in /usr/local/lib
>   libraries mkl,vml,guide not found in /usr/lib
>   NOT AVAILABLE
> 
>   NOT AVAILABLE
> 
> atlas_threads_info:
> Setting PTATLAS=ATLAS
>   libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib
> __main__.atlas_threads_info
>   NOT AVAILABLE
> 
> atlas_info:
>   libraries f77blas,cblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries f77blas,cblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
> __main__.atlas_info
>   FOUND:
>     libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = f77
>     include_dirs = ['/usr/include']
> 
> customize GnuFCompiler
> Found executable /usr/bin/g77
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler using config
>   FOUND:
>     libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = f77
>     define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
>     include_dirs = ['/usr/include']
> 
> lapack_atlas_info:
>   libraries lapack_atlas,f77blas,cblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries lapack_atlas,f77blas,cblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
> __main__.lapack_atlas_info
>   FOUND:
>     libraries = ['lapack', 'lapack_atlas', 'f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = f77
>     include_dirs = ['/usr/include']
> 
> umfpack_info:
>   libraries umfpack not found in /usr/local/lib
>   libraries umfpack not found in /usr/lib
>   NOT AVAILABLE
> 
> _pkg_config_info:
> Found executable /usr/bin/pkg-config
>   NOT AVAILABLE
> 
> lapack_atlas_threads_info:
> Setting PTATLAS=ATLAS
>   libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/local/lib
>   libraries lapack_atlas not found in /usr/local/lib
>   libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
>   libraries lapack_atlas not found in /usr/lib/atlas
>   libraries lapack_atlas,ptf77blas,ptcblas,atlas not found in /usr/lib
> __main__.lapack_atlas_threads_info
>   NOT AVAILABLE
> 
> x11_info:
>   libraries X11 not found in /usr/lib64
>   NOT AVAILABLE
> 
> blas_info:
>   libraries blas not found in /usr/local/lib
>   FOUND:
>     libraries = ['blas']
>     library_dirs = ['/usr/lib']
>     language = f77
> 
> fftw_info:
>   libraries fftw3 not found in /usr/local/lib
>   FOUND:
>     libraries = ['fftw3']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW3_H', None)]
>     include_dirs = ['/usr/include']
> 
> f2py_info:
>   FOUND:
>     sources = ['/usr/lib/python2.5/site-packages/numpy/f2py/src/fortranobject.c']
>     include_dirs = ['/usr/lib/python2.5/site-packages/numpy/f2py/src']
> 
> gdk_pixbuf_xlib_2_info:
>   NOT AVAILABLE
> 
> dfftw_threads_info:
>   libraries drfftw_threads,dfftw_threads not found in /usr/local/lib
>   libraries drfftw_threads,dfftw_threads not found in /usr/lib
>   dfftw threads not found
>   NOT AVAILABLE
> 
> atlas_blas_info:
>   libraries f77blas,cblas,atlas not found in /usr/local/lib
>   libraries f77blas,cblas,atlas not found in /usr/lib/atlas
>   FOUND:
>     libraries = ['f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = c
>     include_dirs = ['/usr/include']
> 
> fftw3_info:
>   libraries fftw3 not found in /usr/local/lib
>   FOUND:
>     libraries = ['fftw3']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW3_H', None)]
>     include_dirs = ['/usr/include']
> 
> blas_opt_info:
> blas_mkl_info:
>   libraries mkl,vml,guide not found in /usr/local/lib
>   libraries mkl,vml,guide not found in /usr/lib
>   NOT AVAILABLE
> 
> atlas_blas_threads_info:
> Setting PTATLAS=ATLAS
>   libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib/atlas
>   libraries ptf77blas,ptcblas,atlas not found in /usr/lib
>   NOT AVAILABLE
> 
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler using config
>   FOUND:
>     libraries = ['f77blas', 'cblas', 'atlas']
>     library_dirs = ['/usr/lib']
>     language = c
>     define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
>     include_dirs = ['/usr/include']
> 
> sfftw_info:
>   libraries srfftw,sfftw not found in /usr/local/lib
>   libraries srfftw,sfftw not found in /usr/lib
>   sfftw not found
>   NOT AVAILABLE
> 
> xft_info:
>   NOT AVAILABLE
> 
> fft_opt_info:
> djbfft_info:
>   NOT AVAILABLE
> 
>   FOUND:
>     libraries = ['fftw3']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW3_H', None)]
>     include_dirs = ['/usr/include']
> 
> gdk_x11_2_info:
>   NOT AVAILABLE
> 
> agg2_info:
>   NOT AVAILABLE
> 
> numarray_info:
>   NOT AVAILABLE
> 
> blas_src_info:
>   NOT AVAILABLE
> 
> fftw2_info:
>   libraries rfftw,fftw not found in /usr/local/lib
>   FOUND:
>     libraries = ['rfftw', 'fftw']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW_H', None)]
>     include_dirs = ['/usr/include']
> 
> fftw_threads_info:
>   libraries rfftw_threads,fftw_threads not found in /usr/local/lib
>   FOUND:
>     libraries = ['rfftw_threads', 'fftw_threads']
>     library_dirs = ['/usr/lib']
>     define_macros = [('SCIPY_FFTW_THREADS_H', None)]
>     include_dirs = ['/usr/include']
> 
> _numpy_info:
>   FOUND:
>     define_macros = [('NUMERIC_VERSION', '"\\"24.2\\""'), ('NUMERIC', None)]
>     include_dirs = ['/usr/include/python2.5']
> 
> wx_info:
> Could not locate executable wx-config
> File not found: None. Cannot determine wx info.
>   NOT AVAILABLE
> 
> gdk_info:
>   NOT AVAILABLE
> 
> gtkp_x11_2_info:
>   NOT AVAILABLE
> 
> sfftw_threads_info:
>   libraries srfftw_threads,sfftw_threads not found in /usr/local/lib
>   libraries srfftw_threads,sfftw_threads not found in /usr/lib
>   sfftw threads not found
>   NOT AVAILABLE
> 
> boost_python_info:
>   NOT AVAILABLE
> 
> freetype2_info:
>   NOT AVAILABLE
> 
> gdk_2_info:
>   NOT AVAILABLE
> 
> dfftw_info:
>   libraries drfftw,dfftw not found in /usr/local/lib
>   libraries drfftw,dfftw not found in /usr/lib
>   dfftw not found
>   NOT AVAILABLE
> 
> lapack_src_info:
>   NOT AVAILABLE
> 
> gtkp_2_info:
>   NOT AVAILABLE
> 
> gdk_pixbuf_2_info:
>   NOT AVAILABLE
> 
> amd_info:
>   libraries amd not found in /usr/local/lib
>   libraries amd not found in /usr/lib
>   NOT AVAILABLE
> 
> Numeric_info:
>   FOUND:
>     define_macros = [('NUMERIC_VERSION', '"\\"24.2\\""'), ('NUMERIC', None)]
>     include_dirs = ['/usr/include/python2.5']
> 
> numerix_info:
> numpy_info:
>   FOUND:
>     define_macros = [('NUMPY_VERSION', '"\\"1.0.4\\""'), ('NUMPY', None)]
> 
>   FOUND:
>     define_macros = [('NUMPY_VERSION', '"\\"1.0.4\\""'), ('NUMPY', None)]
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From timmichelsen at gmx-topmail.de  Mon Nov 26 09:29:45 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Mon, 26 Nov 2007 14:29:45 +0000 (UTC)
Subject: [SciPy-user] Installation of time series failed on windows
Message-ID: <loom.20071126T142633-428@post.gmane.org>

Hello,
I tried to install the time-series package from the sandbox on my windows.
It failed while maskarray was installed nicely.

May someone please help me here?

Here's a log from IPython when running the setup.py install

Ignoring attempt to set 'name' (from 'timeseries' to 'timeseries.lib')
Warning: Assuming default configuration
(.\tests/{setup_tests,setup}.py was not
found)
Appending timeseries.tests configuration to timeseries
Ignoring attempt to set 'name' (from 'timeseries' to 'timeseries.tests')
running install
running build
running config_cc
unifing config_cc, config, build_clib, build_ext, build commands
--compiler options
running config_fc
unifing config_fc, config, build_clib, build_ext, build commands
--fcompiler options
running build_src
building extension "timeseries.cseries" sources
running build_py
copying .\ipython_log.py -> build\lib.win32-2.5\timeseries
running build_ext
No module named msvccompiler in numpy.distutils, trying from distutils..

<type 'exceptions.SystemExit'>: error:
Python was built with Visual Studio 2003;
extensions must be built with a compiler than can generate compatible binaries.
Visual Studio 2003 was not found on this system. If you have Cygwin installed,
you can try compiling with MingW32, by passing "-c mingw32" to setup.py.
WARNING: Failure executing file: <setup.py>

Thanks.
Tim



From Alexander.Dietz at astro.cf.ac.uk  Mon Nov 26 12:05:42 2007
From: Alexander.Dietz at astro.cf.ac.uk (Alexander Dietz)
Date: Mon, 26 Nov 2007 17:05:42 +0000
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
Message-ID: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>

Hi,

I am running FC5 (I think, how to figure out) and I followed the
installation instructions for FC5. The last step, however, fails with the
following error messages:

> python setup.py install
...
...
g77:f77: scipy/fftpack/dfftpack/zfftb.f
g77:f77: scipy/fftpack/dfftpack/dsinti.f
g77:f77: scipy/fftpack/dfftpack/dfftb.f
g77:f77: scipy/fftpack/dfftpack/zfftb1.f
/tmp/ccuB5p4W.s: Assembler messages:
/tmp/ccuB5p4W.s:589: Error: suffix or operands invalid for `movd'
/tmp/ccuB5p4W.s:2948: Error: suffix or operands invalid for `movd'
/tmp/ccuB5p4W.s: Assembler messages:
/tmp/ccuB5p4W.s:589: Error: suffix or operands invalid for `movd'
/tmp/ccuB5p4W.s:2948: Error: suffix or operands invalid for `movd'
error: Command "/usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2
-funroll-loops -march=i686 -mmmx -msse2 -msse -fomit-frame-pointer
-malign-double -c -c scipy/fftpack/dfftpack/zfftb1.f -o build/temp.linux-
i686-2.4/scipy/fftpack/dfftpack/zfftb1.o" failed with exit status 1

Any ideas how to solve this issue?

Cheers
  Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071126/68239337/attachment.html>

From mattknox_ca at hotmail.com  Mon Nov 26 13:07:12 2007
From: mattknox_ca at hotmail.com (Matt Knox)
Date: Mon, 26 Nov 2007 18:07:12 +0000 (UTC)
Subject: [SciPy-user] Installation of time series failed on windows
References: <loom.20071126T142633-428@post.gmane.org>
Message-ID: <loom.20071126T180540-87@post.gmane.org>

Hi Tim,

I just added a section to the wiki about compiling extension modules on 
Windows. Please give the following a try: 
http://www.scipy.org/Cookbook/CompilingExtensionsOnWindowsWithMinGW

If you are still having trouble after that, let me know.

- Matt







From david at ar.media.kyoto-u.ac.jp  Mon Nov 26 22:19:35 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 27 Nov 2007 12:19:35 +0900
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
In-Reply-To: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>
References: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>
Message-ID: <474B8CC7.7010108@ar.media.kyoto-u.ac.jp>

Alexander Dietz wrote:
> Hi,
>
> I am running FC5 (I think, how to figure out) and I followed the 
> installation instructions for FC5. The last step, however, fails with 
> the following error messages:
>
> > python setup.py install
> ...
> ...
> g77:f77: scipy/fftpack/dfftpack/zfftb.f
> g77:f77: scipy/fftpack/dfftpack/dsinti.f
> g77:f77: scipy/fftpack/dfftpack/dfftb.f
> g77:f77: scipy/fftpack/dfftpack/zfftb1.f
> /tmp/ccuB5p4W.s: Assembler messages:
> /tmp/ccuB5p4W.s:589: Error: suffix or operands invalid for `movd'
> /tmp/ccuB5p4W.s:2948: Error: suffix or operands invalid for `movd'
> /tmp/ccuB5p4W.s: Assembler messages:
> /tmp/ccuB5p4W.s:589: Error: suffix or operands invalid for `movd'
> /tmp/ccuB5p4W.s:2948: Error: suffix or operands invalid for `movd'
> error: Command "/usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2 
> -funroll-loops -march=i686 -mmmx -msse2 -msse -fomit-frame-pointer 
> -malign-double -c -c scipy/fftpack/dfftpack/zfftb1.f -o 
> build/temp.linux- i686-2.4/scipy/fftpack/dfftpack/zfftb1.o" failed 
> with exit status 1
>
> Any ideas how to solve this issue?
>
> Cheers
>   Alex
>
>   
What's your machine ? More exactly, could you give us the result of the 
following command (to get the exact characteristics of your CPU): cat 
/proc/cpuinfo ?

cheers,

David


From strawman at astraw.com  Mon Nov 26 22:34:16 2007
From: strawman at astraw.com (Andrew Straw)
Date: Mon, 26 Nov 2007 19:34:16 -0800
Subject: [SciPy-user] pdf of multivariate normal
In-Reply-To: <fh4u8o$6v7$1@ger.gmane.org>
References: <fh4pm3$pvn$1@ger.gmane.org>	<1194717156.583578.152820@o3g2000hsb.googlegroups.com>
	<fh4u8o$6v7$1@ger.gmane.org>
Message-ID: <474B9038.9030708@astraw.com>

John Reid wrote:
> LB wrote:
>>>>> info(numpy.random.multivariate_normal)
>> Return an array containing multivariate normally distributed random
>> numbers
> 
> Thanks but I don't want random variates, I would like to calculate the 
> probability density function.

I have the same question myself. I found one implementation:
scikits.learn.machine.em.densities.gauss_den()

Is there a version in scipy (or numpy)? After searching, I think not,
but it does seems this belongs in there -- perhaps in scipy.stats. Any
reason beyond the usual (lack of time) why it's not there already?

-Andrew


From Alexander.Dietz at astro.cf.ac.uk  Tue Nov 27 03:59:41 2007
From: Alexander.Dietz at astro.cf.ac.uk (Alexander Dietz)
Date: Tue, 27 Nov 2007 08:59:41 +0000
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
In-Reply-To: <474B8CC7.7010108@ar.media.kyoto-u.ac.jp>
References: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>
	<474B8CC7.7010108@ar.media.kyoto-u.ac.jp>
Message-ID: <9cf809a00711270059w117804faj7a03710d8e5fc208@mail.gmail.com>

Hi,

see the content of /proc/cpuinfo below.

Cheers
  Alex

processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 14
model name      : Genuine Intel(R) CPU           T2300  @ 1.66GHz
stepping        : 8
cpu MHz         : 1000.000
cache size      : 2048 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 2
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx constant_tsc pni
monitor est tm2 xtpr
bogomips        : 3335.11

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 6
model           : 14
model name      : Genuine Intel(R) CPU           T2300  @ 1.66GHz
stepping        : 8
cpu MHz         : 1000.000
cache size      : 2048 KB
physical id     : 0
siblings        : 2
core id         : 1
cpu cores       : 2
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx constant_tsc pni
monitor est tm2 xtpr
bogomips        : 3329.30


On Nov 27, 2007 3:19 AM, David Cournapeau <david at ar.media.kyoto-u.ac.jp>
wrote:

> Alexander Dietz wrote:
> > Hi,
> >
> > I am running FC5 (I think, how to figure out) and I followed the
> > installation instructions for FC5. The last step, however, fails with
> > the following error messages:
> >
> > > python setup.py install
> > ...
> > ...
> > g77:f77: scipy/fftpack/dfftpack/zfftb.f
> > g77:f77: scipy/fftpack/dfftpack/dsinti.f
> > g77:f77: scipy/fftpack/dfftpack/dfftb.f
> > g77:f77: scipy/fftpack/dfftpack/zfftb1.f
> > /tmp/ccuB5p4W.s: Assembler messages:
> > /tmp/ccuB5p4W.s:589: Error: suffix or operands invalid for `movd'
> > /tmp/ccuB5p4W.s:2948: Error: suffix or operands invalid for `movd'
> > /tmp/ccuB5p4W.s: Assembler messages:
> > /tmp/ccuB5p4W.s:589: Error: suffix or operands invalid for `movd'
> > /tmp/ccuB5p4W.s:2948: Error: suffix or operands invalid for `movd'
> > error: Command "/usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2
> > -funroll-loops -march=i686 -mmmx -msse2 -msse -fomit-frame-pointer
> > -malign-double -c -c scipy/fftpack/dfftpack/zfftb1.f -o
> > build/temp.linux- i686-2.4/scipy/fftpack/dfftpack/zfftb1.o" failed
> > with exit status 1
> >
> > Any ideas how to solve this issue?
> >
> > Cheers
> >   Alex
> >
> >
> What's your machine ? More exactly, could you give us the result of the
> following command (to get the exact characteristics of your CPU): cat
> /proc/cpuinfo ?
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071127/f5ddd962/attachment.html>

From stefan at sun.ac.za  Tue Nov 27 04:11:41 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Tue, 27 Nov 2007 11:11:41 +0200
Subject: [SciPy-user] pdf of multivariate normal
In-Reply-To: <474B9038.9030708@astraw.com>
References: <fh4pm3$pvn$1@ger.gmane.org>
	<1194717156.583578.152820@o3g2000hsb.googlegroups.com>
	<fh4u8o$6v7$1@ger.gmane.org> <474B9038.9030708@astraw.com>
Message-ID: <20071127091141.GD29033@mentat.za.net>

On Mon, Nov 26, 2007 at 07:34:16PM -0800, Andrew Straw wrote:
> I have the same question myself. I found one implementation:
> scikits.learn.machine.em.densities.gauss_den()
> 
> Is there a version in scipy (or numpy)? After searching, I think not,
> but it does seems this belongs in there -- perhaps in scipy.stats. Any
> reason beyond the usual (lack of time) why it's not there already?

Isn't learn.machine.em.densities.gauss_den used to estimate parameters
for Gaussian mixture models?  For a multivariate Gaussian, the mean
and covariance can be calculated directly.

Regards
St?fan


From david at ar.media.kyoto-u.ac.jp  Tue Nov 27 04:06:50 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 27 Nov 2007 18:06:50 +0900
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
In-Reply-To: <9cf809a00711270059w117804faj7a03710d8e5fc208@mail.gmail.com>
References: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>	<474B8CC7.7010108@ar.media.kyoto-u.ac.jp>
	<9cf809a00711270059w117804faj7a03710d8e5fc208@mail.gmail.com>
Message-ID: <474BDE2A.5000404@ar.media.kyoto-u.ac.jp>

Alexander Dietz wrote:
> Hi,
>
> see the content of /proc/cpuinfo below.
>
Ok, your CPU supports sse2 instruction set, so this is not the problem. 
Since you use an old FC version, it may be an old g77 bug. Could you 
give use the version of g77 (g77 -v). Also, you could try to see which 
flag make it fails, for example, does the following:

/usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2 -funroll-loops 
-march=i686 -mmmx -msse -fomit-frame-pointer -malign-double -c 
scipy/fftpack/dfftpack/zfftb1.f

(I suspect the -msse2 to be a problem with old g77; several problems 
related to -msse2 and movd were fixed in gcc 3.3).

cheers,

David


From david at ar.media.kyoto-u.ac.jp  Tue Nov 27 04:18:00 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 27 Nov 2007 18:18:00 +0900
Subject: [SciPy-user] pdf of multivariate normal
In-Reply-To: <20071127091141.GD29033@mentat.za.net>
References: <fh4pm3$pvn$1@ger.gmane.org>	<1194717156.583578.152820@o3g2000hsb.googlegroups.com>	<fh4u8o$6v7$1@ger.gmane.org>
	<474B9038.9030708@astraw.com>
	<20071127091141.GD29033@mentat.za.net>
Message-ID: <474BE0C8.3080206@ar.media.kyoto-u.ac.jp>

Stefan van der Walt wrote:
> On Mon, Nov 26, 2007 at 07:34:16PM -0800, Andrew Straw wrote:
>> I have the same question myself. I found one implementation:
>> scikits.learn.machine.em.densities.gauss_den()
>>
>> Is there a version in scipy (or numpy)? After searching, I think not,
>> but it does seems this belongs in there -- perhaps in scipy.stats. Any
>> reason beyond the usual (lack of time) why it's not there already?
>
> Isn't learn.machine.em.densities.gauss_den used to estimate parameters
> for Gaussian mixture models?  For a multivariate Gaussian, the mean
> and covariance can be calculated directly.
No, it is used to compute the pdf of a multivariate Gaussian, which is 
then used to compute the pdf of mixtures. AFAIK, there is no 
multivariate facilities in scipy.stats, or did I missed them ? Providing 
the same facilities as in scipy.stats.distribution for multi-variate 
would be quite a lot of coding, though I guess. For example, for cdf 
alone, it is not so easy to compute them generally (but the code is 
already there for kde if I am right, in the mvn extension).

David


From Alexander.Dietz at astro.cf.ac.uk  Tue Nov 27 04:39:01 2007
From: Alexander.Dietz at astro.cf.ac.uk (Alexander Dietz)
Date: Tue, 27 Nov 2007 09:39:01 +0000
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
In-Reply-To: <474BDE2A.5000404@ar.media.kyoto-u.ac.jp>
References: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>
	<474B8CC7.7010108@ar.media.kyoto-u.ac.jp>
	<9cf809a00711270059w117804faj7a03710d8e5fc208@mail.gmail.com>
	<474BDE2A.5000404@ar.media.kyoto-u.ac.jp>
Message-ID: <9cf809a00711270139h3e5ee58bk62f8af9f6322f08c@mail.gmail.com>

Hi,

On Nov 27, 2007 9:06 AM, David Cournapeau <david at ar.media.kyoto-u.ac.jp>
wrote:

> Alexander Dietz wrote:
> > Hi,
> >
> > see the content of /proc/cpuinfo below.
> >
> Ok, your CPU supports sse2 instruction set, so this is not the problem.
> Since you use an old FC version, it may be an old g77 bug. Could you
> give use the version of g77 (g77 -v).


Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/3.2.3/specs
Configured with: ../configure --prefix=/usr --mandir=/usr/share/man
--infodir=/usr/share/info --enable-shared --enable-threads=posix
--disable-checking --with-system-zlib --enable-__cxa_atexit
--enable-languages=c,c++,f77 --disable-libgcj --host=i386-redhat-linux
Thread model: posix
gcc version 3.2.3 20030502 (Red Hat Linux 3.2.3-55.fc5)



> Also, you could try to see which
> flag make it fails, for example, does the following:
>
> /usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2 -funroll-loops
> -march=i686 -mmmx -msse -fomit-frame-pointer -malign-double -c
> scipy/fftpack/dfftpack/zfftb1.f
>

That worked actually...


Cheers
  Alex



> (I suspect the -msse2 to be a problem with old g77; several problems
> related to -msse2 and movd were fixed in gcc 3.3).
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071127/aec96e20/attachment.html>

From david at ar.media.kyoto-u.ac.jp  Tue Nov 27 04:43:15 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 27 Nov 2007 18:43:15 +0900
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
In-Reply-To: <9cf809a00711270139h3e5ee58bk62f8af9f6322f08c@mail.gmail.com>
References: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>	<474B8CC7.7010108@ar.media.kyoto-u.ac.jp>	<9cf809a00711270059w117804faj7a03710d8e5fc208@mail.gmail.com>	<474BDE2A.5000404@ar.media.kyoto-u.ac.jp>
	<9cf809a00711270139h3e5ee58bk62f8af9f6322f08c@mail.gmail.com>
Message-ID: <474BE6B3.9030709@ar.media.kyoto-u.ac.jp>

Alexander Dietz wrote:
>  
>
>     Also, you could try to see which
>     flag make it fails, for example, does the following:
>
>     /usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2 -funroll-loops
>     -march=i686 -mmmx -msse -fomit-frame-pointer -malign-double -c
>     scipy/fftpack/dfftpack/zfftb1.f
>
>
> That worked actually... 
Yes, so the problem is your gcc version. Maybe we could disable adding 
-msse2 altogether for gcc < 3.3, instead of enabling it for gcc > 3.2.2 
(look at numpy/distutils/fcompiler/gnu.py, and look for -msse2; if Pearu 
or David are reading this, maybe they know a better solution ?).

Note that FC5 is old. You should consider upgrading (FC 5 is not 
maintained anymore) if you can,

cheers,

David



From david at ar.media.kyoto-u.ac.jp  Tue Nov 27 04:46:43 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 27 Nov 2007 18:46:43 +0900
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
In-Reply-To: <474BE6B3.9030709@ar.media.kyoto-u.ac.jp>
References: <9cf809a00711260905o3c0d14behed14925157c0be9b@mail.gmail.com>	<474B8CC7.7010108@ar.media.kyoto-u.ac.jp>	<9cf809a00711270059w117804faj7a03710d8e5fc208@mail.gmail.com>	<474BDE2A.5000404@ar.media.kyoto-u.ac.jp>	<9cf809a00711270139h3e5ee58bk62f8af9f6322f08c@mail.gmail.com>
	<474BE6B3.9030709@ar.media.kyoto-u.ac.jp>
Message-ID: <474BE783.2020707@ar.media.kyoto-u.ac.jp>

David Cournapeau wrote:
> Alexander Dietz wrote:
>   
>>  
>>
>>     Also, you could try to see which
>>     flag make it fails, for example, does the following:
>>
>>     /usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2 -funroll-loops
>>     -march=i686 -mmmx -msse -fomit-frame-pointer -malign-double -c
>>     scipy/fftpack/dfftpack/zfftb1.f
>>
>>
>> That worked actually... 
>>     
> Yes, so the problem is your gcc version. Maybe we could disable adding 
> -msse2 altogether for gcc < 3.3, instead of enabling it for gcc > 3.2.2 
> (look at numpy/distutils/fcompiler/gnu.py, and look for -msse2; if Pearu 
> or David are reading this, maybe they know a better solution ?).
>
> Note that FC5 is old. You should consider upgrading (FC 5 is not 
> maintained anymore) if you can,
I forgot: another possibility is to use gfortran instead of g77, but you 
will have to recompile everything (you cannot really mix  g77 and 
gfortran code together), including BLAS/LAPACK. For FC6, I remember that 
gfortran is the default ABI, so this is anyway the best choice, but I am 
not sure for FC5.

David


From cookedm at physics.mcmaster.ca  Tue Nov 27 08:57:37 2007
From: cookedm at physics.mcmaster.ca (David M. Cooke)
Date: Tue, 27 Nov 2007 08:57:37 -0500
Subject: [SciPy-user] install numpy/scipy with Intel Compilers
In-Reply-To: <47471912.3050300@imag.fr>
References: <47471912.3050300@imag.fr>
Message-ID: <F14293F6-0734-4725-9F97-31446B8F5C9D@physics.mcmaster.ca>

On Nov 23, 2007, at 13:16 , Laurence Viry wrote:

> Hi!
>
> I try to install numpy 1.0.4 and scipy 0.6.0 with INTEL C/C++ and
> Fortran compilers and SCSL scientific library with python2.4.4 on
> Itanium with linux OS. (Altix SGI with SUSE linux)
>
> I didn't success to configure the construct of build_ext for scipy  
> with
> the fortran INTEL compiler (--fcompiler=intele)
> Thank for your help

Hi Laurence,

I'm not familiar with Intel's compilers (even more, running them on  
Itaniums). I'm guessing that numpy.distutils isn't able to tell  
whether your compiler is intele or not. From numpy/distutils/fcompiler/ 
intel.py, the command run to find the version is

$ ifort -FI -V -c dummy.f -o dummy.o

where dummy.f is just a file with a dummy subroutine in it. Could you  
do this and give us the output?

-- 
|>|\/|<
/------------------------------------------------------------------\
|David M. Cooke              http://arbutus.physics.mcmaster.ca/dmc/
|cookedm at physics.mcmaster.ca



From rex at nosyntax.net  Tue Nov 27 10:40:00 2007
From: rex at nosyntax.net (rex)
Date: Tue, 27 Nov 2007 07:40:00 -0800
Subject: [SciPy-user] install numpy/scipy with Intel Compilers
In-Reply-To: <F14293F6-0734-4725-9F97-31446B8F5C9D@physics.mcmaster.ca>
References: <47471912.3050300@imag.fr>
	<F14293F6-0734-4725-9F97-31446B8F5C9D@physics.mcmaster.ca>
Message-ID: <20071127154000.GE5720@nosyntax.net>

David M. Cooke <cookedm at physics.mcmaster.ca> [2007-11-27 05:54]:
>I'm guessing that numpy.distutils isn't able to tell  
>whether your compiler is intele or not. From numpy/distutils/fcompiler/ 
>intel.py, the command run to find the version is
>
>$ ifort -FI -V -c dummy.f -o dummy.o
>
>where dummy.f is just a file with a dummy subroutine in it. Could you  
>do this and give us the output?

Hello David,

No Itanium here, but this is the output of the 32-bit version.
==============================================================================
$ ifort -FI -V -c foo.f -o foo.o
Intel(R) Fortran Compiler for applications running on IA-32, Version 10.1    Build 20070913 Package ID: l_fc_p_10.1.008
Copyright (C) 1985-2007 Intel Corporation.  All rights reserved.
FOR NON-COMMERCIAL USE ONLY

 Intel Fortran 10.0-2024
================================================================================


There's a distutils problem using Intel's MKL in NumPy-1.04 and
NumPy SVN 4510 (and below). 

In MKL 9.1 & 10.0, mkl_lapack32 & mkl_lapack64 no longer
exist. They were replaced by /32/mkl_lapack and
/64/mkl_lapack. As a result,
numpy/distutils/system_info.py
needs to be changed:

class lapack_mkl_info(mkl_info):

    def calc_info(self):
        mkl = get_info('mkl')
        if not mkl:
            return
        if sys.platform == 'win32':
            lapack_libs = self.get_libs('lapack_libs',['mkl_lapack'])
        else:
            #lapack_libs = self.get_libs('lapack_libs',['mkl_lapack32','mkl_lapack64'])
            lapack_libs = self.get_libs('lapack_libs',['mkl_lapack'])


If this change is not made NumPy builds, but the MKL is not
used. I reported this problem earlier on the NumPy list. 

-rex


From Laurence.Viry at imag.fr  Tue Nov 27 11:32:25 2007
From: Laurence.Viry at imag.fr (Laurence Viry)
Date: Tue, 27 Nov 2007 17:32:25 +0100
Subject: [SciPy-user] install numpy/scipy with Intel Compilers
In-Reply-To: <F14293F6-0734-4725-9F97-31446B8F5C9D@physics.mcmaster.ca>
References: <47471912.3050300@imag.fr>
	<F14293F6-0734-4725-9F97-31446B8F5C9D@physics.mcmaster.ca>
Message-ID: <474C4699.7020905@imag.fr>

D
>
> Hi Laurence,
>
> I'm not familiar with Intel's compilers (even more, running them on  
> Itaniums). I'm guessing that numpy.distutils isn't able to tell  
> whether your compiler is intele or not. From numpy/distutils/fcompiler/ 
> intel.py, the command run to find the version is
>
> $ ifort -FI -V -c dummy.f -o dummy.o
>
> where dummy.f is just a file with a dummy subroutine in it. Could you  
> do this and give us the output?
>
>   
* the ouput is the following,
---------------------------------
Intel(R) Fortran Itanium(R) Compiler for Itanium(R)-based applications
Version 9.0    Build 20050430 Package ID: l_fc_p_9.0.021
Copyright (C) 1985-2005 Intel Corporation.  All rights reserved.
FOR NON-COMMERCIAL USE ONLY
----------------------------------
I observed before, it was not the same

in intel.py
description = 'Intel Fortran Compiler for Itanium apps'

I tried to replace by :
description = 'Intel(R) Fortran Itanium(R) Compiler for Itanium(R) apps'

it's ok for numpy but I had stil the same error for scipy

* I don't use MKL bul SCSL library, it is the optimized scientific 
library on our computrer,  I give you the ouput of the config command
blas_info:
  FOUND:
    libraries = ['scs']
    library_dirs = ['/usr/lib']
    language = f77

  FOUND:
    libraries = ['scs']
    library_dirs = ['/usr/lib']
    define_macros = [('NO_ATLAS_INFO', 1)]
    language = f77

lapack_info:
  FOUND:
    libraries = ['scs']
    library_dirs = ['/usr/lib']
    language = f77

  FOUND:
    libraries = ['scs', 'scs']
    library_dirs = ['/usr/lib']
    define_macros = [('NO_ATLAS_INFO', 1)]
    language = f77

Thank for your help

-- 
Laurence Viry

CRIP UJF - Projet MIRAGE ( http://mirage.imag.fr )
Laboratoire de Mod?lisation et Calcul - IMAG

tel: 04 76 63 56 34
fax: 04 76 63 12 63
e-mail: Laurence.Viry at imag.fr



From odonnems at yahoo.com  Wed Nov 28 14:47:00 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Wed, 28 Nov 2007 11:47:00 -0800 (PST)
Subject: [SciPy-user] weave inline compile error
Message-ID: <943333.38163.qm@web58012.mail.re3.yahoo.com>

I have been trying to use scipy.weave.inline on Windows XP via cygwin. Ultimately, I need to use weave.inline for compiling c++ code nested in python program. However, I have been extremely unsuccessful at getting this to work. I recently posted some errors on this site, which are subject under "weave and compiler install help". Basically, I have not been able to find out anything else, but let me summarize:


First, I am using these versions, but note I have tried to use the most recent stable release of python enthought:
I created a distutils.cfg file specifying 
[build_ext]
compiler=mingw32

python-2.4.1.msi
scipy-0.6.0.win32-py2.4.exe
Numeric-24.2.win32-py2.4.exe
numpy-1.04.win32-py2.4.exe
wxPython2.8-win32-unicode-2.8.6.1-py24.exe  and wxPython2.8-win32-docs-demos-2.8.6.1.exe
As well as a couple other packages
cygwin: setup.exe (current)
mingw-install-20060210.tar (MSYS) and MinGW-5.1.3.exe (I read online that these do not work with 2.4.1 so I am focusing with cygwin)


An example of weave that I am trying to get to work is:
a = 1
weave.inline('printf("%d\\n",a);',['a'], verbose=2, type_converters=converters.blitz, compiler='gcc') #I have tried numerous versions of this


The error seems to be in the code written in the numpy\distutils\exec_command.py:
s,o = exec_command(cmd)

The best I can tell, is that the code is trying to execute the above command wich is failing to write output to s,o.  These  variables are then written to files, which cannot be read. The program jumps around between exec_command and _exec_command. I don't know python well enough to know what is causing the error. It appears to me that the code is being compiled, but this one line is not working. I also tried running the script written by exec_command.py directly (commented out the os.remove() in exec_command.py, but this did not work. It did help me identify that the problem is with s and o variables.

Does anyone have any thoughts on what I should try to do. I appreciate any suggestions and ideas one can provide.

Thanks!
Michael


Here is an example of my errors:
running build_ext
running build_src
building extension "sc_552cccf5dbf4f6eadd273cdcbd5860525" sources
customize Mingw32CCompiler
customize Mingw32CCompiler using build_ext
customize Mingw32CCompiler
customize Mingw32CCompiler using build_ext
building 'sc_552cccf5dbf4f6eadd273cdcbd5860525' extension
compiling C++ sources
C compiler: g++ -mno-cygwin -O2 -Wall

compile options: '-IC:\Python24\lib\site-packages\scipy\weave -IC:\Python24\lib\site-packages\scipy\weave\scxx -IC:\Python24\lib\site-packages\numpy\core\include -IC:\Python24\include -IC:\Python24\PC -c'
g++ -mno-cygwin -O2 -Wall -IC:\Python24\lib\site-packages\scipy\weave -IC:\Python24\lib\site-packages\scipy\weave\scxx -IC:\Python24\lib\site-packages\numpy\core\include -IC:\Python24\include -IC:\Python24\PC -c c:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860525.cpp -o c:\temp\Michael\python24_intermediate\compiler_c8350d870e6c54e8f29dd7094c2bfb45\Release\temp\michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860525.o
Traceback (most recent call last):
  File "C:\Python24\Lib\site-packages\pythonwin\pywin\framework\scriptutils.py", line 310, in RunScript
    exec codeObject in __main__.__dict__
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 251, in ?
    main()
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 205, in main
    weave.inline('printf("%d\\n",a);',['a'], verbose=2, type_converters=converters.blitz)  #, compiler = 'gcc', verbpse=2, type_converters=converters.blitz, auto_downcast=0) #'msvc' or 'gcc' or 'mingw32'
  File "C:\Python24\Lib\site-packages\scipy\weave\inline_tools.py", line 338, in inline
    auto_downcast = auto_downcast,
  File "C:\Python24\Lib\site-packages\scipy\weave\inline_tools.py", line 447, in compile_function
    verbose=verbose, **kw)
  File "C:\Python24\Lib\site-packages\scipy\weave\ext_tools.py", line 365, in compile
    verbose = verbose, **kw)
  File "C:\Python24\Lib\site-packages\scipy\weave\build_tools.py", line 269, in build_extension
    setup(name = module_name, ext_modules = [ext],verbose=verb)
  File "C:\Python24\Lib\site-packages\numpy\distutils\core.py", line 176, in setup
    return old_setup(**new_attr)
  File "C:\Python24\lib\distutils\core.py", line 159, in setup
    raise SystemExit, error
CompileError: error: c:\temp\tmpgnzrbr\buunmb.txt: No such file or directory






      ____________________________________________________________________________________
Be a better pen pal. 
Text or chat with friends inside Yahoo! Mail. See how.  http://overview.mail.yahoo.com/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071128/d4243ba8/attachment.html>

From williamhpurcell at gmail.com  Wed Nov 28 19:36:23 2007
From: williamhpurcell at gmail.com (William Purcell)
Date: Wed, 28 Nov 2007 18:36:23 -0600
Subject: [SciPy-user] Scipy in XP on VMWare Workstation
Message-ID: <474E0987.7050000@gmail.com>

I am trying to run scipy in XP on VMWare Workstation. The host OS is 
Ubuntu.  When I attempt to start scipy I get the error:
____________________________________________________________________________

C:\Documents and Settings\Administrator>C:\Python25\python.exe 
C:\Python25\scrip

ts\ipython -pylab -p scipy

Traceback (most recent call last):

File "C:\Python25\scripts\ipython", line 27, in <module>

IPython.Shell.start().mainloop()

File "C:\Python25\Lib\site-packages\IPython\Shell.py", line 1111, in start

return shell(user_ns = user_ns)

File "C:\Python25\Lib\site-packages\IPython\Shell.py", line 1008, in 
__init__

shell_class=MatplotlibShell)

File "C:\Python25\Lib\site-packages\IPython\Shell.py", line 74, in __init__

debug=debug,shell_class=shell_class)

File "C:\Python25\Lib\site-packages\IPython\ipmaker.py", line 95, in 
make_IPyt

hon

embedded=embedded,**kw)

File "C:\Python25\Lib\site-packages\IPython\Shell.py", line 562, in __init__

user_ns,b2 = self._matplotlib_config(name,user_ns)

File "C:\Python25\Lib\site-packages\IPython\Shell.py", line 463, in 
_matplotli

b_config

from matplotlib import backends

File "C:\Python25\Lib\site-packages\matplotlib\backends\__init__.py", 
line 55,

in <module>

new_figure_manager, draw_if_interactive, show = pylab_setup()

File "C:\Python25\Lib\site-packages\matplotlib\backends\__init__.py", 
line 24,

in pylab_setup

globals(),locals(),[backend_name])

File 
"C:\Python25\Lib\site-packages\matplotlib\backends\backend_tkagg.py", lin

e 8, in <module>

import tkagg # Paint image to Tk photo blitter extension

File "C:\Python25\Lib\site-packages\matplotlib\backends\tkagg.py", line 
1, in

<module>

import _tkagg

ImportError: DLL load failed: The specified module could not be found.

_____________________________________________________________________________________

Has anyone been successful in doing this or have any suggested fixes.

Thanks,
Bill


From hbabcock at mac.com  Wed Nov 28 21:45:26 2007
From: hbabcock at mac.com (Hazen Babcock)
Date: Wed, 28 Nov 2007 21:45:26 -0500
Subject: [SciPy-user] spe file reader (Roper Scientific / Winview)
Message-ID: <3E7F13DD-576C-4390-B130-90AB4310DC37@mac.com>


Hello,

Does anyone know of a module for reading (and possibly writing) Roper  
Scientific / Winview .spe formatted files?

thanks,
-Hazen



From vincefn at users.sourceforge.net  Thu Nov 29 03:37:51 2007
From: vincefn at users.sourceforge.net (Favre-Nicolin Vincent)
Date: Thu, 29 Nov 2007 09:37:51 +0100
Subject: [SciPy-user] spe file reader (Roper Scientific / Winview)
In-Reply-To: <3E7F13DD-576C-4390-B130-90AB4310DC37@mac.com>
References: <3E7F13DD-576C-4390-B130-90AB4310DC37@mac.com>
Message-ID: <200711290937.51949.vincefn@users.sourceforge.net>

	Hi,

> Does anyone know of a module for reading (and possibly writing) Roper
> Scientific / Winview .spe formatted files?

   The following file should provide reading - I'm using it to read ccd images 
from princeton camers. No writing though. I'm not sure how many variants of 
the format this can read so be careful !
   And it only reads the image data, not the associated info (date etc...).

	Vincent
-- 
Vincent Favre-Nicolin
Universit? Joseph Fourier
http://v.favrenicolin.free.fr
ObjCryst & Fox : http://objcryst.sourceforge.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: winview.py
Type: application/x-python
Size: 4822 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071129/391047cd/attachment.bin>

From discerptor at gmail.com  Thu Nov 29 13:17:31 2007
From: discerptor at gmail.com (Joshua Lippai)
Date: Thu, 29 Nov 2007 10:17:31 -0800
Subject: [SciPy-user] Error in scipy.test(10) even after compiling
	successfully- Mac OS X 10.4.11 Apple GCC 4.0.1
Message-ID: <9911419a0711291017r22d48580q2c1652855d253c86@mail.gmail.com>

Hello all,

After building scipy, I ran the scipy tests and got this error in check_integer:

======================================================================
ERROR: check_integer (scipy.io.tests.test_array_import.TestReadArray)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/io/tests/test_array_import.py",
line 55, in check_integer
    from scipy import stats
  File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/stats/__init__.py",
line 7, in <module>
    from stats import *
  File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/stats/stats.py",
line 191, in <module>
    import scipy.special as special
  File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/__init__.py",
line 8, in <module>
    from basic import *
  File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/basic.py",
line 8, in <module>
    from _cephes import *
ImportError: dlopen(/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/_cephes.so,
2): Symbol not found: __gfortran_pow_r8_i4
  Referenced from:
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/_cephes.so
  Expected in: dynamic lookup

It would appear it's not finding a symbol for gfrotran in one of the
files, and this in turn makes it impossible to import _cephes. How
would I go about remedying this?


From david at ar.media.kyoto-u.ac.jp  Thu Nov 29 20:58:19 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 30 Nov 2007 10:58:19 +0900
Subject: [SciPy-user] Error in scipy.test(10) even after
 compiling	successfully- Mac OS X 10.4.11 Apple GCC 4.0.1
In-Reply-To: <9911419a0711291017r22d48580q2c1652855d253c86@mail.gmail.com>
References: <9911419a0711291017r22d48580q2c1652855d253c86@mail.gmail.com>
Message-ID: <474F6E3B.1020209@ar.media.kyoto-u.ac.jp>

Joshua Lippai wrote:
> Hello all,
>
> After building scipy, I ran the scipy tests and got this error in check_integer:
>
> ======================================================================
> ERROR: check_integer (scipy.io.tests.test_array_import.TestReadArray)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/io/tests/test_array_import.py",
> line 55, in check_integer
>     from scipy import stats
>   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/stats/__init__.py",
> line 7, in <module>
>     from stats import *
>   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/stats/stats.py",
> line 191, in <module>
>     import scipy.special as special
>   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/__init__.py",
> line 8, in <module>
>     from basic import *
>   File "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/basic.py",
> line 8, in <module>
>     from _cephes import *
> ImportError: dlopen(/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/_cephes.so,
> 2): Symbol not found: __gfortran_pow_r8_i4
>   Referenced from:
> /Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/special/_cephes.so
>   Expected in: dynamic lookup
>
> It would appear it's not finding a symbol for gfrotran in one of the
> files, and this in turn makes it impossible to import _cephes. How
> would I go about remedying this?
>   
This symbol is in the gfortran runtime. We need the complete build log 
to be sure about the exact problem, but I expect some mismatch between 
fortran compilers (do you have several fortran compilers on your machine ?).

cheers,

David



From lbolla at gmail.com  Fri Nov 30 05:12:39 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Fri, 30 Nov 2007 11:12:39 +0100
Subject: [SciPy-user] problems installing scipy on windows with cygwin
Message-ID: <80c99e790711300212q4974fb80j5aa5d6401dcb2430@mail.gmail.com>

I'm having problems installing scipy on my windows machine (Quad
2.3GHzIntel Pentium processors) using cygwin.

Here is the error message:

-------------------------

<many lines here>

/cygdrive/c/Documents and Settings/bollalo001/My Documents/archive/scipy-
0.6.0/build/src.cygwin-1.5.24-i686-2.5/build/src.cygwin-1.5.24-i686-2.5/scipy/lib/blas/fblasmodule.c:474:
undefined reference to `_csrot_'
/cygdrive/c/Documents and Settings/bollalo001/My Documents/archive/scipy-
0.6.0/build/src.cygwin-1.5.24-i686-2.5/build/src.cygwin-1.5.24-i686-2.5/scipy/lib/blas/fblasmodule.c:487:
undefined reference to `_zdrot_'

collect2: ld returned 1 exit status
error: Command "/usr/bin/g77 -g -Wall -g -Wall -shared build/temp.cygwin-
1.5.24-i686-2.5/build/src.cygwin-1.5.24-i686-2.5/build/src.cygwin-
1.5.24-i686-2.5/scipy/lib/blas/fblasmodule.o build/temp.cygwin-
1.5.24-i686-2.5/build/src.cygwin-1.5.24-i686-2.5/fortranobject.o
build/temp.cygwin-1.5.24-i686-2.5/build/src.cygwin-1.5.24-i686-2.5/scipy/lib/blas/fblaswrap.o
build/temp.cygwin-1.5.24-i686-2.5/build/src.cygwin-1.5.24-i686-2.5
/build/src.cygwin-1.5.24-i686-2.5/scipy/lib/blas/fblas-f2pywrappers.o-L/usr/lib
-L/usr/lib/gcc/i686-pc-cygwin/3.4.4 -L/usr/lib/python2.5/config
-Lbuild/temp.cygwin-1.5.24-i686-2.5 -lblas -lpython2.5 -lg2c -o
build/lib.cygwin-1.5.24-i686-2.5/scipy/lib/blas/fblas.dll" failed with exit
status 1
-------------------------



It seems like a LAPACK problem. I tried with both the cygwin provided
blas/lapack and with the complete BLAS/LAPACK from netlib.

what do you suggest?

thank you!

L.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071130/c3f12200/attachment.html>

From lbolla at gmail.com  Fri Nov 30 08:33:59 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Fri, 30 Nov 2007 14:33:59 +0100
Subject: [SciPy-user] enthough python: installing sandbox packages
Message-ID: <80c99e790711300533s36c07903n20fc0d3e4c67bc98@mail.gmail.com>

Hi all,
is there a simple way to install sandbox packages (e.g. arpack) in a
Enthough python using tools like easy_install.exe or enstaller.exe?
Thanks in advance,
L.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071130/0916a508/attachment.html>

From Giovanni.Samaey at cs.kuleuven.be  Fri Nov 30 12:26:50 2007
From: Giovanni.Samaey at cs.kuleuven.be (Giovanni Samaey)
Date: Fri, 30 Nov 2007 18:26:50 +0100
Subject: [SciPy-user] lil_matrix
Message-ID: <A53FBE8F-6299-46FF-9AEB-1D4BF9CB4CE8@cs.kuleuven.be>

Hello all,

I have a basic question concerning lil_matrix slice assignments.
The following code does not behave as intended :

import scipy.sparse as S

A = S.lil_matrix((5,5))
A.setdiag(ones(5))
B = S.lil_matrix((6,6))
B[:5,:5]=A
B.getnnz()

This returns 0 instead of the expected 5.  Am I using the code  
incorrectly ?

Giovanni


From jh at physics.ucf.edu  Fri Nov 30 15:00:29 2007
From: jh at physics.ucf.edu (Joe Harrington)
Date: Fri, 30 Nov 2007 15:00:29 -0500
Subject: [SciPy-user] surface plotting
Message-ID: <1196452829.6047.387.camel@glup.physics.ucf.edu>

Well, I was hoping for a lighter solution, but I gave it a one-hour try.
It didn't help that when I did the link to the user docs was broken
(it's fixed now).  I got so far as installing mayavi2 and trying to
figure it out without the doc, but ran out of time and wasn't able to
figure out surface plotting my existing numpy array.  This was for use
in a class and I decided that if I couldn't figure out how to do it in
an hour I probably couldn't expect the students to do it at all.

It would be good to have a cookbook recipe about how to do surface
plotting of a 2D numpy array in mayavi2.

When will mayavi2 be available as a deb file?  This looks like software
that everyone should be able to get in the easiest fashion possible.

Thanks,

--jh--




From jdh2358 at gmail.com  Fri Nov 30 15:07:54 2007
From: jdh2358 at gmail.com (John Hunter)
Date: Fri, 30 Nov 2007 14:07:54 -0600
Subject: [SciPy-user] surface plotting
In-Reply-To: <1196452829.6047.387.camel@glup.physics.ucf.edu>
References: <1196452829.6047.387.camel@glup.physics.ucf.edu>
Message-ID: <88e473830711301207h62f83491x78c5a217179bce41@mail.gmail.com>

On Nov 30, 2007 2:00 PM, Joe Harrington <jh at physics.ucf.edu> wrote:

> It would be good to have a cookbook recipe about how to do surface
> plotting of a 2D numpy array in mayavi2.

For the purposes of your class, I find that a simple imshow or pcolor
of 2D data with a color map is usually good enough to visualize a 2D
array.  Surf doesn't really add much unless you are interactively
panning and zooming, when the movement generates extra 3D visual cues.

JDH


From fredmfp at gmail.com  Fri Nov 30 15:12:10 2007
From: fredmfp at gmail.com (fred)
Date: Fri, 30 Nov 2007 21:12:10 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <1196452829.6047.387.camel@glup.physics.ucf.edu>
References: <1196452829.6047.387.camel@glup.physics.ucf.edu>
Message-ID: <47506E9A.4090705@gmail.com>

Joe Harrington a ?crit :
> It would be good to have a cookbook recipe about how to do surface
> plotting of a 2D numpy array in mayavi2.
>   
http://www.scipy.org/Cookbook/MayaVi/Surf
http://www.scipy.org/Cookbook/MayaVi/mlab
> When will mayavi2 be available as a deb file?
It's coming... :-)


HTH.

Cheers,

-- 
http://scipy.org/FredericPetit



From robert.kern at gmail.com  Fri Nov 30 15:19:58 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 30 Nov 2007 14:19:58 -0600
Subject: [SciPy-user] enthough python: installing sandbox packages
In-Reply-To: <80c99e790711300533s36c07903n20fc0d3e4c67bc98@mail.gmail.com>
References: <80c99e790711300533s36c07903n20fc0d3e4c67bc98@mail.gmail.com>
Message-ID: <4750706E.70308@gmail.com>

lorenzo bolla wrote:
> Hi all,
> is there a simple way to install sandbox packages (e.g. arpack) in a
> Enthough python using tools like easy_install.exe or enstaller.exe?

Unfortunately, there isn't. The way the sandbox is set up, the sandbox packages
are built as part of the monolithic scipy package.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco


From jh at physics.ucf.edu  Fri Nov 30 15:28:25 2007
From: jh at physics.ucf.edu (Joe Harrington)
Date: Fri, 30 Nov 2007 15:28:25 -0500
Subject: [SciPy-user] surface plotting
Message-ID: <1196454505.6047.418.camel@glup.physics.ucf.edu>

> For the purposes of your class, I find that a simple imshow or pcolor
> of 2D data with a color map is usually good enough to visualize a 2D
> array.  Surf doesn't really add much unless you are interactively
> panning and zooming, when the movement generates extra 3D visual cues.

Mostly we use a combination of imshow and ds9 in astronomy.  There is a
*lot* of zooming and panning, and real-time, interactive colormap and
scaling manipulations, and zooming the cursor around and watching what
values go under the cursor.  If there's a way to do any of those in
matplotlib it would be good to know how, but I didn't see it in the
manual.  The numdisplay interface to ds9 is pretty good, but not
perfect, as it interpolates the image before sending it to ds9, which is
a throwback to an interface defined in the days of yore.

But, the particular application for which surface really stands out is
looking at the relative brightnesses of stars.  Yeah, you can get an
idea of it with a colormap, but green is not three times red, and even
shades of gray are hard to divide by eye.  Differing heights are easy to
do by eye.

Within that category, the one thing I simply can't do well with an image
display is look at bad pixels.  To find these, and find if you've killed
them all, you really need to do an edge-on look at a surface plot.  With
images that are several k by several k, a single pixel just gets
swallowed up in an image display.  Its spike does not get swallowed up.
Also, a good 2D display will examine the histogram and scale to, say 98%
of the histogram so that spurious spikes in the data with values of
10**12 don't make it impossible to see the stars.

For this particular application, surface really is the only good way
I've found.

As I said in an earlier post, the matplotlib surface looks fine, except
that it is too slow.  I'm not sure if that's because the other methods
I've used are in compiled code or if there is some better algorithm.
But, I've been using surface plots for ~20 years, and the computers back
then were slower than current ones by more than the difference between
compiled and python code, I think.  The IRAF package is open-source and
has such an algorithm, though it's in a bowdlerized version of FORTRAN.

--jh--




From rmay at ou.edu  Fri Nov 30 16:56:26 2007
From: rmay at ou.edu (Ryan May)
Date: Fri, 30 Nov 2007 15:56:26 -0600
Subject: [SciPy-user] 1D Correlation on 3D array
Message-ID: <4750870A.5080101@ou.edu>

Hi,

I was curious if anyone has a fast non-looping way to calculate 1D
correlations on a N-dim array (my case is specifically 3).

What i have right now is rather slow:

import numpy as N
from scipy.signal import correlate

ts_data = N.ones((360,1201,17), dtype=N.complex64)
for index in N.ndindex(ts_data.shape[:-1]):
    x = ts_data[index]
    R = correlate(x, x.conj())

I've looked at just passing in the whole array, but that seems to want
to do full N-dim correlation.

Thanks,

Ryan

-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma