From prabhu at aero.iitb.ac.in  Sat Dec  1 11:48:05 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Sat, 01 Dec 2007 22:18:05 +0530
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: <47519045.4040800@aero.iitb.ac.in>

Joe Harrington wrote:
> 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.

Sorry about the mess with installation instructions etc.  I'm trying to 
get some help regarding this but I've been too busy otherwise to do much 
about it -- end of semester here.  I'm hoping to try and get something 
done about it over the next couple of weeks.

One of the easiest ways to get things installed without eggs right now 
is to first install the dependencies -- wxPython 2.6, python-vtk and 
numpy and then grab the tarballs here:

  http://code.enthought.com/downloads/source/ets2.6/

Untar them and build them each

  cd enthought.traits
  ./build.sh
  vi/emacs install.sh
  ./install.sh

for all the tarballs and you should have mayavi installed.  The script 
is broken in that it does not install the mayavi executable though -- 
need to fix that.

> It would be good to have a cookbook recipe about how to do surface
> plotting of a 2D numpy array in mayavi2.

mayavi.tools.mlab.surf should help here.

> 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.

Ondrej Certik has been very helpful in making preliminary debs that we 
hope will soon get into debian.  You can find a older versions here:

  http://debian.certik.cz/

I sent Ondrej patches to fix some of the issues but don't see new 
versions yet.  Ondrej did say he was busy...


cheers,
prabhu



From prabhu at aero.iitb.ac.in  Sat Dec  1 11:54:42 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Sat, 01 Dec 2007 22:24:42 +0530
Subject: [SciPy-user] surface plotting
In-Reply-To: <47506E9A.4090705@gmail.com>
References: <1196452829.6047.387.camel@glup.physics.ucf.edu>
	<47506E9A.4090705@gmail.com>
Message-ID: <475191D2.3050305@aero.iitb.ac.in>

fred wrote:
> 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

Unfortunately, these still use tvtk's mlab which is lower level and not 
as powerful as mayavi's new mlab.

The best docs for that are here:

  https://svn.enthought.com/enthought/attachment/wiki/MayaVi/mlab.pdf


cheers,
prabhu


From lbolla at gmail.com  Sat Dec  1 14:46:13 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Sat, 1 Dec 2007 20:46:13 +0100
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: <80c99e790712011146l1db7827dje3af20c5f01b635f@mail.gmail.com>

I had the same problem in Linux.
But in Linux there are no .pyd files.

I noticed that there were two shared libraries: sparsetools.so and
_sparsetools.so.
The first one was much smaller than the second, so I moved it to a backup
directory.

Then, everything worked fine.

hth,
L.


On Nov 18, 2007 11:50 AM, youngsu park <youngsu999 at gmail.com> wrote:

> 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 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/20071201/5a3ae160/attachment.html>

From wizzard028wise at gmail.com  Sat Dec  1 15:37:42 2007
From: wizzard028wise at gmail.com (Dorian)
Date: Sat, 1 Dec 2007 21:37:42 +0100
Subject: [SciPy-user] Algorithm of Viterbi
Message-ID: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>

Hi all,

Is there any tools about the algorithm of Viterbi using scypi ?.
Any link on a clear tutorial will be very helpful.

Thank in advance
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071201/3abb020b/attachment.html>

From emanuelez at gmail.com  Sat Dec  1 15:57:29 2007
From: emanuelez at gmail.com (Emanuele Zattin)
Date: Sat, 1 Dec 2007 21:57:29 +0100
Subject: [SciPy-user] Algorithm of Viterbi
In-Reply-To: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
References: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
Message-ID: <d7b942cd0712011257j7ab3a726h72d2d421c26312be@mail.gmail.com>

Well... it seems that Wikipedia has some Python code about it. It
shouldn't be too difficult to modify it in order to use Numpy.

Just my 0.02$

On Dec 1, 2007 9:37 PM, Dorian <wizzard028wise at gmail.com> wrote:
> Hi all,
>
> Is there any tools about the algorithm of Viterbi using scypi ?.
> Any link on a clear tutorial will be very helpful.
>
> Thank in advance
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>



-- 
Emanuele Zattin
---------------------------------------------------
-I don't have to know an answer. I don't feel frightened by not
knowing things; by being lost in a mysterious universe without any
purpose ? which is the way it really is, as far as I can tell,
possibly. It doesn't frighten me.- Richard Feynman


From jelleferinga at gmail.com  Sat Dec  1 17:27:22 2007
From: jelleferinga at gmail.com (jelle)
Date: Sat, 1 Dec 2007 22:27:22 +0000 (UTC)
Subject: [SciPy-user] Algorithm of Viterbi
References: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
	<d7b942cd0712011257j7ab3a726h72d2d421c26312be@mail.gmail.com>
Message-ID: <loom.20071201T222603-854@post.gmane.org>

check out this article, that should the thing your looking for

http://www.biais.org/blog/index.php/?q=viterbi



From wizzard028wise at gmail.com  Sat Dec  1 18:41:21 2007
From: wizzard028wise at gmail.com (Dorian)
Date: Sun, 2 Dec 2007 00:41:21 +0100
Subject: [SciPy-user] Algorithm of Viterbi
In-Reply-To: <loom.20071201T222603-854@post.gmane.org>
References: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
	<d7b942cd0712011257j7ab3a726h72d2d421c26312be@mail.gmail.com>
	<loom.20071201T222603-854@post.gmane.org>
Message-ID: <674a602a0712011541l4e17d87aodeecae99f001c563@mail.gmail.com>

thank you very much

On 01/12/2007, jelle <jelleferinga at gmail.com> wrote:
>
> check out this article, that should the thing your looking for
>
> http://www.biais.org/blog/index.php/?q=viterbi
>
> _______________________________________________
> 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/20071202/6652a923/attachment.html>

From lorenzo.isella at gmail.com  Sun Dec  2 10:24:50 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Sun, 02 Dec 2007 16:24:50 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <mailman.11.1196532009.17180.scipy-user@scipy.org>
References: <mailman.11.1196532009.17180.scipy-user@scipy.org>
Message-ID: <4752CE42.60406@gmail.com>

Dear All,
I refer to the text quoted below. I am interested in data representation 
and 3D plots in particular (something matplotlib is not very suitable for).

>One of the easiest ways to get things installed without eggs right now 
>is to first install the dependencies -- wxPython 2.6, python-vtk and 
>numpy and then grab the tarballs here:

  http://code.enthought.com/downloads/source/ets2.6/
>
>Untar them and build them each

 > cd enthought.traits
 > ./build.sh
 > vi/emacs install.sh
 > ./install.sh

>for all the tarballs and you should have mayavi installed.  The script 
>is broken in that it does not install the mayavi executable though -- 
>need to fix that.

In case it matters, I am running Debian testing on my box.
I tried first using eggs (with little success) and then the tarballs + 
./build.sh + ./install.sh.
Now, I am having 2 problems:
(1) if I try to re-run some old codes of mine(relying on SciPy), I get 
tons of these warnings:
/usr/lib/python2.4/site-packages/scipy/misc/__init__.py:25: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/linalg/__init__.py:32: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/optimize/__init__.py:17: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/ndimage/__init__.py:40: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/sparse/__init__.py:9: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/io/__init__.py:20: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/lib/__init__.py:5: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/linsolve/umfpack/__init__.py:7: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/linsolve/__init__.py:13: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/interpolate/__init__.py:15: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/special/__init__.py:22: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/stats/__init__.py:15: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/fftpack/__init__.py:21: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/integrate/__init__.py:16: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/signal/__init__.py:17: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
/usr/lib/python2.4/site-packages/scipy/maxentropy/__init__.py:12: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test

Basically, ScipyTest is now, in some sense, NumpyTest. Other than that, 
the codes seem to run fine, but I hope I have not harmed my system. Does 
it have anything to do with the different versions of Python I have 
installed? Should I be worried to end up with a broken Python? That 
would be a disaster for me. What is giving rise to this new warning? Any 
suggestions here are welcome.
(2)This is less critical to me: when I try running the code snippet on
 http://www.scipy.org/Cookbook/MayaVi/mlab

import scipy

# prepare some interesting function:
def f(x, y):
    return 3.0*scipy.sin(x*y+1e-4)/(x*y+1e-4)

x = scipy.arange(-7., 7.05, 0.1)
y = scipy.arange(-5., 5.05, 0.1)

# 3D visualization of f:
from enthought.tvtk.tools import mlab
fig = mlab.figure()
s = mlab.SurfRegular(x, y, f)
fig.add(s)


this is what I get:

/usr/lib/python2.4/site-packages/scipy/misc/__init__.py:25: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test

(python:13574): Gtk-CRITICAL **: gtk_widget_set_colormap: assertion 
`!GTK_WIDGET_REALIZED (widget)' failed


and no window opens. Any idea of what is going on? BTW, I actually tried 
replacing  ScipyTest with NumpyTest  in 
/usr/lib/python2.4/site-packages/scipy/misc/__init__.py:25, but then the 
code crashed badly (can give more info if needed). So, one error and the 
same warning as before. The other example I want to discuss is
(http://www.scipy.org/Cookbook/MayaVi/Surf)


import numpy
def f(x, y):
return numpy.sin(x*y)/(x*y)
x = numpy.arange(-7., 7.05, 0.1)
y = numpy.arange(-5., 5.05, 0.05)
from enthought.tvtk.tools import mlab
s = mlab.SurfRegular(x, y, f)
from enthought.mayavi.sources.vtk_data_source import VTKDataSource
d = VTKDataSource()
d.data = s.data
mayavi.add_source(d)
from enthought.mayavi.filters.warp_scalar import WarpScalar
w = WarpScalar()
mayavi.add_filter(w)
from enthought.mayavi.modules.outline import Outline
from enthought.mayavi.modules.surface import Surface
o = Outline()
s = Surface()
mayavi.add_module(o)
mayavi.add_module(s)


but again,  this is what I get:

/usr/lib/python2.4/site-packages/scipy/misc/__init__.py:25: 
DeprecationWarning: ScipyTest is now called NumpyTest; please update 
your code
  test = ScipyTest().test
So far so good
Traceback (most recent call last):
  File "3D-enthought.py", line 13, in ?
    mayavi.add_source(d)
NameError: name 'mayavi' is not defined

So, what is going on here? Mayavi is actually installed on my system...

Sorry for the long post. My main interest is point (1), i.e. to be sure 
that my python installation is still OK (and know what to do in case it 
is not). As to point (2), in the end of the day, these modules will make 
it into debian packages, won't they? So at some point installing them 
should be straightforward (at least for me; I am not very good at 
handling source and installing myself).
Many thanks

Lorenzo












From gael.varoquaux at normalesup.org  Sun Dec  2 11:40:17 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 2 Dec 2007 17:40:17 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <4752CE42.60406@gmail.com>
References: <mailman.11.1196532009.17180.scipy-user@scipy.org>
	<4752CE42.60406@gmail.com>
Message-ID: <20071202164017.GE15280@clipper.ens.fr>

On Sun, Dec 02, 2007 at 04:24:50PM +0100, Lorenzo Isella wrote:
> I refer to the text quoted below. I am interested in data representation 
> and 3D plots in particular (something matplotlib is not very suitable for).

Sure. Tell us a bit more about your usecase. Do you want to produce
figures for print/publication, do you want to interact with your data, or
do you want simply to do some 3D plotting in a program of yours. You can
do all this with Mayavi, but the solutions will depend a bit of your usecase.


>The other example I want to discuss is
> (http://www.scipy.org/Cookbook/MayaVi/Surf)


> import numpy
> def f(x, y):
> return numpy.sin(x*y)/(x*y)
> x = numpy.arange(-7., 7.05, 0.1)
> y = numpy.arange(-5., 5.05, 0.05)
> from enthought.tvtk.tools import mlab
> s = mlab.SurfRegular(x, y, f)
> from enthought.mayavi.sources.vtk_data_source import VTKDataSource
> d = VTKDataSource()
> d.data = s.data
> mayavi.add_source(d)
> from enthought.mayavi.filters.warp_scalar import WarpScalar
> w = WarpScalar()
> mayavi.add_filter(w)
> from enthought.mayavi.modules.outline import Outline
> from enthought.mayavi.modules.surface import Surface
> o = Outline()
> s = Surface()
> mayavi.add_module(o)
> mayavi.add_module(s)


> but again,  this is what I get:

> /usr/lib/python2.4/site-packages/scipy/misc/__init__.py:25: 
> DeprecationWarning: ScipyTest is now called NumpyTest; please update 
> your code
>   test = ScipyTest().test
> So far so good
> Traceback (most recent call last):
>   File "3D-enthought.py", line 13, in ?
>     mayavi.add_source(d)
> NameError: name 'mayavi' is not defined

> So, what is going on here? Mayavi is actually installed on my system...

This code has to be run through mayavi. It is suited only for interactive
visualisation or producing figures, and not to run as part of a Python
program. You have to run it in mayavi, either by running "mayavi2 -n -x
script.py", loading it through the menu (File -> Open File), and pressing
Ctrl+R, or entering "execfile('script.py') in the python shell.

> Sorry for the long post. 

No problem, that's what mailing lists are for.

> My main interest is point (1), i.e. to be sure that my python
> installation is still OK (and know what to do in case it is not). 

I think it is OK. What happened is that scipy got updated, and these are
just harmless, albeit annoying, warnings.

> As to point (2), in the end of the day, these modules will make 
> it into debian packages, won't they? 

Yes, we are working on them. I had to finish my PhD thesis, so I was
lagging behind on my work on packaging and Co. I am mostly done with
that (!!), so things will start to move a bit faster soon.

> So at some point installing them should be straightforward (at least
> for me; I am not very good at handling source and installing myself).

Yes, we are working on that. We know it is very important and will focus
effort on ease of instal.

Cheers,

Ga?l



From gael.varoquaux at normalesup.org  Sun Dec  2 11:49:44 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 2 Dec 2007 17:49:44 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <475191D2.3050305@aero.iitb.ac.in>
References: <1196452829.6047.387.camel@glup.physics.ucf.edu>
	<47506E9A.4090705@gmail.com> <475191D2.3050305@aero.iitb.ac.in>
Message-ID: <20071202164944.GF15280@clipper.ens.fr>

On Sat, Dec 01, 2007 at 10:24:42PM +0530, Prabhu Ramachandran wrote:
> Unfortunately, these still use tvtk's mlab which is lower level and not 
> as powerful as mayavi's new mlab.

> The best docs for that are here:

>   https://svn.enthought.com/enthought/attachment/wiki/MayaVi/mlab.pdf


Quickly, you can have a feeling of how to plot surface using mlab by
folliwing these steps:

run "ipython -wthread"

execute the following python code:

+++++++++++++++++++++++++++++++++++++++
In [1]: from enthought.mayavi.tools import mlab as M
Set Envisage to use the workbench UI: True
Set Envisage to use the workbench UI: True

In [2]: from scipy import *

In [3]: X, Y = mgrid[-1:1:50j, -1:1:50j]

In [4]: Z = cos(4*(X**2+Y**2))

In [5]: M.surf(X, Y, Z)
+++++++++++++++++++++++++++++++++++++++


This will bring up a Mayavi2 window with the surface plotted. The mlab
API is similar to matplotlib's pylab, so you should feel quite familiar
with it. It is still young and I haven't had time to put as much work as
I would like in it. Documentation is lacking, but the docstrings are
quite good. Just reading "help(M)" in the example above should hopefully
give you an idea of the possbilities. The docs pointed to by Prabhu
should also be helpful.

HTH,

Ga?l


From william.ratcliff at gmail.com  Sun Dec  2 12:56:47 2007
From: william.ratcliff at gmail.com (william ratcliff)
Date: Sun, 2 Dec 2007 12:56:47 -0500
Subject: [SciPy-user] surface plotting
In-Reply-To: <20071202164944.GF15280@clipper.ens.fr>
References: <1196452829.6047.387.camel@glup.physics.ucf.edu>
	<47506E9A.4090705@gmail.com> <475191D2.3050305@aero.iitb.ac.in>
	<20071202164944.GF15280@clipper.ens.fr>
Message-ID: <827183970712020956j4652119m658d2901abcac5ee@mail.gmail.com>

Can Mayavi2 be called from within a wxpython application for 3D plotting?
At one point I did this with tvtk (using pyface, etc), but not with mayavi.
How different is the mlab within tvtk from that in mayavi2?

Thanks,
William


On Dec 2, 2007 11:49 AM, Gael Varoquaux <gael.varoquaux at normalesup.org>
wrote:

> On Sat, Dec 01, 2007 at 10:24:42PM +0530, Prabhu Ramachandran wrote:
> > Unfortunately, these still use tvtk's mlab which is lower level and not
> > as powerful as mayavi's new mlab.
>
> > The best docs for that are here:
>
> >   https://svn.enthought.com/enthought/attachment/wiki/MayaVi/mlab.pdf
>
>
> Quickly, you can have a feeling of how to plot surface using mlab by
> folliwing these steps:
>
> run "ipython -wthread"
>
> execute the following python code:
>
> +++++++++++++++++++++++++++++++++++++++
> In [1]: from enthought.mayavi.tools import mlab as M
> Set Envisage to use the workbench UI: True
> Set Envisage to use the workbench UI: True
>
> In [2]: from scipy import *
>
> In [3]: X, Y = mgrid[-1:1:50j, -1:1:50j]
>
> In [4]: Z = cos(4*(X**2+Y**2))
>
> In [5]: M.surf(X, Y, Z)
> +++++++++++++++++++++++++++++++++++++++
>
>
> This will bring up a Mayavi2 window with the surface plotted. The mlab
> API is similar to matplotlib's pylab, so you should feel quite familiar
> with it. It is still young and I haven't had time to put as much work as
> I would like in it. Documentation is lacking, but the docstrings are
> quite good. Just reading "help(M)" in the example above should hopefully
> give you an idea of the possbilities. The docs pointed to by Prabhu
> should also be helpful.
>
> HTH,
>
> Ga?l
> _______________________________________________
> 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/20071202/5b184547/attachment.html>

From gael.varoquaux at normalesup.org  Sun Dec  2 13:07:48 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 2 Dec 2007 19:07:48 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <827183970712020956j4652119m658d2901abcac5ee@mail.gmail.com>
References: <1196452829.6047.387.camel@glup.physics.ucf.edu>
	<47506E9A.4090705@gmail.com> <475191D2.3050305@aero.iitb.ac.in>
	<20071202164944.GF15280@clipper.ens.fr>
	<827183970712020956j4652119m658d2901abcac5ee@mail.gmail.com>
Message-ID: <20071202180748.GL15280@clipper.ens.fr>

On Sun, Dec 02, 2007 at 12:56:47PM -0500, william ratcliff wrote:
>    Can Mayavi2 be called from within a wxpython application for 3D plotting?

Yes. This has been one of the great additions of this summer. Have a look
at
https://svn.enthought.com/enthought/browser/branches/enthought.mayavi_2.0.3a1/examples/standalone.py
for a demo of how mayavi is integrated in a traits application. As traits
relies on WxPython (althought, there is now a Qt front end), it is trivial to
plug this in a WxPython applictation (ask if you need help).

>    At one point I did this with tvtk (using pyface, etc), but not with
>    mayavi.  How different is the mlab within tvtk from that in mayavi2?

I more and more think of the mayavi engine as an abstraction over tvtk
that allows one to think in terms of data and visualisation rather than
in terms of VTK objects. The mayavi2 application is a GUI around the
mayavi2 engine. The example above exposes the engine without the GUI,
allowing you to use its nice API and features in you own app. Mlab is a
simplied API for Mayavi that tries to mimic pylab/matlab. The calls are
really simple and easy to learn. However, due to a design error on my
side, mlab cannot be used in a standalone way, ie to integrate in an
existing WxPython app. This will be changed, just give me a month or two
(not that there is a huge amount of work, but that I am very busy).

It is much easier to add features to mayavi's mlab than to tvtk's.
Moreover, with mayavi's mlab you can pop up the UI to edit the pipeline
in your own app, you also have the engine API that you can access, and
this is great to update/modify a visualization. I think the big
difference between the two is the presence of the mayavi2 engine, which
give a nice abstraction on the visulization. As a result, mayavi's mlab
already has more features than tvtk's, altought some features of tvtk's
mlab are missing (mainly the ability to run in standalone).

HTH,

Ga?l


From odonnems at yahoo.com  Mon Dec  3 00:15:13 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Sun, 2 Dec 2007 21:15:13 -0800 (PST)
Subject: [SciPy-user] Weave inline errors further insight
Message-ID: <808613.94047.qm@web58004.mail.re3.yahoo.com>

I have posted a couple questions regarding the use of weave.inline. Because I have not received any feed back I thought I might provide some more clues to the problem that I am having. 

As a refresher, I have the following versions installed:
numpy-1.0.4.win32-py2.4.exe
scipy-0.6.0.win32-py2.4.exe
python-2.4.4.msi
wxPython2.8-win32-unicode-2.8.6.1-py24.exe
pywin32-210.win32-py2.4.exe
I have also installed SDK toolkit and other related modifications (This appears to work)
I also installed MinGW and this appears to work.
MS XP SP2

First problem: There appears to be a problem with numpy/distutils/exec_command.py. I had to comment out lines 67-70 (there about) in order to properly assign the python executable. Once I do this I get the next error.

Second problem: Will not complete the weave.inline. Example of code here:
a = 1
weave.inline('printf("%d\\n",a);',['a'], verbose=2, type_converters=converters.blitz)

The c++ source file is created; however, the object file is not created. This would make me think there is a problem with the comiler but I have tried both SDK and MinGW and they both cause the same error.

Does anyone have thoughts on why this might be happening. I get the same problem if I insall enthought as well. I will cross-post on numpy too.


Thank you for your assistance,
michael (new user)




      ____________________________________________________________________________________
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/20071202/977193bd/attachment.html>

From fperez.net at gmail.com  Mon Dec  3 00:54:14 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Sun, 2 Dec 2007 22:54:14 -0700
Subject: [SciPy-user] Weave inline errors further insight
In-Reply-To: <808613.94047.qm@web58004.mail.re3.yahoo.com>
References: <808613.94047.qm@web58004.mail.re3.yahoo.com>
Message-ID: <db6b5ecc0712022154n5f55b379of2617ddd4b4a4c00@mail.gmail.com>

On Dec 2, 2007 10:15 PM, Michael ODonnell <odonnems at yahoo.com> wrote:
>
> I have posted a couple questions regarding the use of weave.inline. Because
> I have not received any feed back I thought I might provide some more clues
> to the problem that I am having.
>
> As a refresher, I have the following versions installed:
> numpy-1.0.4.win32-py2.4.exe
> scipy-0.6.0.win32-py2.4.exe
> python-2.4.4.msi
> wxPython2.8-win32-unicode-2.8.6.1-py24.exe
> pywin32-210.win32-py2.4.exe
> I have also installed SDK toolkit and other related modifications (This
> appears to work)
> I also installed MinGW and this appears to work.
> MS XP SP2
>
> First problem: There appears to be a problem with
> numpy/distutils/exec_command.py. I had to comment out lines 67-70 (there
> about) in order to properly assign the python executable. Once I do this I
> get the next error.
>
> Second problem: Will not complete the weave.inline. Example of code here:
> a = 1
> weave.inline('printf("%d\\n",a);',['a'], verbose=2,
> type_converters=converters.blitz)
>
> The c++ source file is created; however, the object file is not created.
> This would make me think there is a problem with the comiler but I have
> tried both SDK and MinGW and they both cause the same error.
>
> Does anyone have thoughts on why this might be happening. I get the same
> problem if I insall enthought as well. I will cross-post on numpy too.
>
>
> Thank you for your assistance,
> michael (new user)
>
>  ________________________________
> Be a better pen pal. Text or chat with friends inside Yahoo! Mail. See how.
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


From fperez.net at gmail.com  Mon Dec  3 01:06:00 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Sun, 2 Dec 2007 23:06:00 -0700
Subject: [SciPy-user] Weave inline errors further insight
In-Reply-To: <808613.94047.qm@web58004.mail.re3.yahoo.com>
References: <808613.94047.qm@web58004.mail.re3.yahoo.com>
Message-ID: <db6b5ecc0712022206k19c692a2jbc9fcf1783c230db@mail.gmail.com>

[ sorry for the previous empty repy.  Slow javascript in gmail
swallowed two clicks...]

On Dec 2, 2007 10:15 PM, Michael ODonnell <odonnems at yahoo.com> wrote:
>
> I have posted a couple questions regarding the use of weave.inline. Because
> I have not received any feed back I thought I might provide some more clues
> to the problem that I am having.

Could I please pester you to open a ticket on the weave problems you
may be having, with *actual example code* that fails, including the
explicit (copy/paste) error messages from the code exactly as you ran
it?

This will allow us to specifically address those problems, if we can
reproduce them.

Thanks,

f


From fperez.net at gmail.com  Mon Dec  3 01:09:32 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Sun, 2 Dec 2007 23:09:32 -0700
Subject: [SciPy-user] weave and compiler install help
In-Reply-To: <271728.69580.qm@web58012.mail.re3.yahoo.com>
References: <271728.69580.qm@web58012.mail.re3.yahoo.com>
Message-ID: <db6b5ecc0712022209x713a4866ved5abf572bc9dc95@mail.gmail.com>

On Nov 19, 2007 9:06 AM, Michael ODonnell <odonnems at yahoo.com> wrote:
>
> 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.

> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 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)

There may be windows-specific problems (I'm on linux), but your code
as above simply can't work because you're assuming blitz behavior
(parens on m(i,j) calls) with the blitz call commented out.

On Linux, this version:

##########################################################
import numpy
from scipy import  weave
from scipy.weave import converters


def prod(m, v):
    #C++ version
    nrows, ncolumns = m.shape
    assert v.ndim==1 and ncolumns==v.shape[0],"Shape mismatch in prod"

    res = numpy.zeros(nrows, 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'], verbose=2,
    type_converters=converters.blitz)
    return res

m=numpy.random.rand(3,4)
v = numpy.random.rand(4)
mv = prod(m,v)
mvd = numpy.dot(m,v)
print 'Err:',numpy.linalg.norm(mv-mvd)
#######  EOF  ###############################################

produces errors within floating-point tolerances, as expected:

In [41]: run weaveex.py
Err: 1.11022302463e-16

In [42]: run weaveex.py
Err: 5.55111512313e-17

In [43]: run weaveex.py
Err: 2.77555756156e-17


As I said in my other message, it's still worth filing a ticket on
this with the exact input/output pair that produced the error.  If
it's a core weave bug we'll find it in any platform, and even if it's
platform specific it will ensure that it doesn't get lost in the mists
of mailing lists.

Cheers,

f


From gael.varoquaux at normalesup.org  Mon Dec  3 01:35:27 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 3 Dec 2007 07:35:27 +0100
Subject: [SciPy-user] Weave inline errors further insight
In-Reply-To: <808613.94047.qm@web58004.mail.re3.yahoo.com>
References: <808613.94047.qm@web58004.mail.re3.yahoo.com>
Message-ID: <20071203063527.GD10287@clipper.ens.fr>

On Sun, Dec 02, 2007 at 09:15:13PM -0800, Michael ODonnell wrote:
>    Second problem: Will not complete the weave.inline. Example of code here:
>    a = 1
>    weave.inline('printf("%d\\n",a);',['a'], verbose=2,
>    type_converters=converters.blitz)

>    The c++ source file is created; however, the object file is not created.
>    This would make me think there is a problem with the comiler but I have
>    tried both SDK and MinGW and they both cause the same error.

Hum. Works here (that doesn't help you sorry). That might be a
platform-specific problem, in which case I am at lost.

Ga?l


From discerptor at gmail.com  Mon Dec  3 05:03:16 2007
From: discerptor at gmail.com (Joshua Lippai)
Date: Mon, 3 Dec 2007 02:03:16 -0800
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: <474F6E3B.1020209@ar.media.kyoto-u.ac.jp>
References: <9911419a0711291017r22d48580q2c1652855d253c86@mail.gmail.com>
	<474F6E3B.1020209@ar.media.kyoto-u.ac.jp>
Message-ID: <9911419a0712030203r7d631be8qdd67e2f3acef5038@mail.gmail.com>

On Nov 29, 2007 5:58 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
>
> 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
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>

That sounds about right... looking at the build dialog, certain
commands seem to want to use gnu95 while the others want gfortran.

unifing config_fc, config, build_clib, build_ext, build commands
--fcompiler options
  commands have different --fcompiler options: ['gfortran', 'gnu95'],
using first in list as default

I tried setting the --fcompiler option for all of the commands
mentioned there to gfortran (at least it changed the default from
gnu95 to gfortran like I wanted), but it still gives that message.
Anything else I can do to make this confusion for the compiler go
away?

Josh


From dmitrey.kroshko at scipy.org  Mon Dec  3 05:14:28 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Mon, 03 Dec 2007 12:14:28 +0200
Subject: [SciPy-user] [optimization] BSD-licensed constrained solver for
	NLP/NSP
Message-ID: <4753D704.1090702@scipy.org>

Hi all,
if anyone is interested - here is one more constrained NLP/NSP solver 
(OpenOpt ralg). 1st derivatives df, dc, dh can be handled as well.

Currently it has no difference of providing some constraints (c[i]<=0, 
h[j]=0) or single: max_i,j_(c[i], abs(h[j]))=0.
But in future I intend to add a little bit better constraints handling 
in case of several funcs.

See the NLP example published some days ago with ralg:
http://openopt.blogspot.com/2007/12/constraints-handling-for-nlpnsp-solver.html

Regards, D.


From prabhu at aero.iitb.ac.in  Mon Dec  3 06:48:28 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Mon, 03 Dec 2007 17:18:28 +0530
Subject: [SciPy-user] surface plotting
In-Reply-To: <20071202180748.GL15280@clipper.ens.fr>
References: <1196452829.6047.387.camel@glup.physics.ucf.edu>	<47506E9A.4090705@gmail.com>
	<475191D2.3050305@aero.iitb.ac.in>	<20071202164944.GF15280@clipper.ens.fr>	<827183970712020956j4652119m658d2901abcac5ee@mail.gmail.com>
	<20071202180748.GL15280@clipper.ens.fr>
Message-ID: <4753ED0C.6090406@aero.iitb.ac.in>

Gael Varoquaux wrote:
> On Sun, Dec 02, 2007 at 12:56:47PM -0500, william ratcliff wrote:
>>    Can Mayavi2 be called from within a wxpython application for 3D plotting?
> 
> Yes. This has been one of the great additions of this summer. Have a look
> at
> https://svn.enthought.com/enthought/browser/branches/enthought.mayavi_2.0.3a1/examples/standalone.py
> for a demo of how mayavi is integrated in a traits application. As traits
> relies on WxPython (althought, there is now a Qt front end), it is trivial to
> plug this in a WxPython applictation (ask if you need help).

I wouldn't say it is "trivial" but certainly relatively easy to do.  The 
best place to ask would be on the enthought-dev list.

>>    At one point I did this with tvtk (using pyface, etc), but not with
>>    mayavi.  How different is the mlab within tvtk from that in mayavi2?
[...]
> allowing you to use its nice API and features in you own app. Mlab is a
> simplied API for Mayavi that tries to mimic pylab/matlab. The calls are
> really simple and easy to learn. However, due to a design error on my
> side, mlab cannot be used in a standalone way, ie to integrate in an
> existing WxPython app. This will be changed, just give me a month or two
> (not that there is a huge amount of work, but that I am very busy).

Right, I don't think it should be too hard to port mlab to use the raw 
engine.  There are a few design issues Gael and I will need to sort out 
about how best to do this though.  As Gael said, both of us are 
currently tied up it may take a month or two for this to actually happen.

cheers,
prabhu


From lbolla at gmail.com  Mon Dec  3 07:10:50 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Mon, 3 Dec 2007 13:10:50 +0100
Subject: [SciPy-user] scipy 0.6.0 ImportError: cannot import name cscmux
In-Reply-To: <80c99e790712011146l1db7827dje3af20c5f01b635f@mail.gmail.com>
References: <338357900711180250u1fd08b86ide9a504d664e0cc1@mail.gmail.com>
	<80c99e790712011146l1db7827dje3af20c5f01b635f@mail.gmail.com>
Message-ID: <80c99e790712030410h3375343ejf41bc33375dc8506@mail.gmail.com>

ops. I was wrong.
it still doesn't work...
sorry for the noise.
L.


On 12/1/07, lorenzo bolla <lbolla at gmail.com> wrote:
>
> I had the same problem in Linux.
> But in Linux there are no .pyd files.
>
> I noticed that there were two shared libraries: sparsetools.so and
> _sparsetools.so.
> The first one was much smaller than the second, so I moved it to a backup
> directory.
>
> Then, everything worked fine.
>
> hth,
> L.
>
>
>  On Nov 18, 2007 11:50 AM, youngsu park <youngsu999 at gmail.com> wrote:
>
> >  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 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/20071203/7ea987c1/attachment.html>

From matthieu.brucher at gmail.com  Mon Dec  3 07:31:50 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Mon, 3 Dec 2007 13:31:50 +0100
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: <e76aa17f0712030431x731347b9r3ebd8ef0497f1eaa@mail.gmail.com>

How did you install scipy ? Which Python version ? ... ?

2007/11/18, youngsu park <youngsu999 at 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 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/20071203/de538e05/attachment.html>

From lbolla at gmail.com  Mon Dec  3 07:41:27 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Mon, 3 Dec 2007 13:41:27 +0100
Subject: [SciPy-user] scipy 0.6.0 ImportError: cannot import name cscmux
In-Reply-To: <e76aa17f0712030431x731347b9r3ebd8ef0497f1eaa@mail.gmail.com>
References: <338357900711180250u1fd08b86ide9a504d664e0cc1@mail.gmail.com>
	<e76aa17f0712030431x731347b9r3ebd8ef0497f1eaa@mail.gmail.com>
Message-ID: <80c99e790712030441n50e96096pb7e764303f592178@mail.gmail.com>

 python 2.4
last numpy (from svn)
last scipy (from svn)

the strange thing is: it works on cygwin on a windows box, but not on
ubuntu!


On 12/3/07, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:
>
> How did you install scipy ? Which Python version ? ... ?
>
> 2007/11/18, youngsu park <youngsu999 at 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 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
> _______________________________________________
> 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/20071203/15bd6f4d/attachment.html>

From cimrman3 at ntc.zcu.cz  Mon Dec  3 07:52:42 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Mon, 03 Dec 2007 13:52:42 +0100
Subject: [SciPy-user] scipy 0.6.0 ImportError: cannot import name cscmux
In-Reply-To: <80c99e790712030441n50e96096pb7e764303f592178@mail.gmail.com>
References: <338357900711180250u1fd08b86ide9a504d664e0cc1@mail.gmail.com>	<e76aa17f0712030431x731347b9r3ebd8ef0497f1eaa@mail.gmail.com>
	<80c99e790712030441n50e96096pb7e764303f592178@mail.gmail.com>
Message-ID: <4753FC1A.9080900@ntc.zcu.cz>

The newly added README in scipy/sparse/sparsetools/README.txt says

"""
Before regenerating sparsetools_wrap.cxx with SWIG, ensure that you
are using SWIG Version 1.3.33 (released on Nov. 23 2007) or newer.
You can check your version with: 'swig -version'

The wrappers are generated with:
    swig -c++ -python multigridtools.i
"""

maybe your problems are related to this?

btw.

Out[1]: <unittest._TextTestResult run=1809 errors=0 failures=0>
on Gentoo linux, updated just now.

r.

lorenzo bolla wrote:
>  python 2.4
> last numpy (from svn)
> last scipy (from svn)
> 
> the strange thing is: it works on cygwin on a windows box, but not on
> ubuntu!
> 
> 
> On 12/3/07, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:
>> How did you install scipy ? Which Python version ? ... ?
>>
>> 2007/11/18, youngsu park <youngsu999 at 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 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
>> _______________________________________________
>> 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 stefan at sun.ac.za  Mon Dec  3 10:28:24 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Mon, 3 Dec 2007 17:28:24 +0200
Subject: [SciPy-user] Algorithm of Viterbi
In-Reply-To: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
References: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
Message-ID: <20071203152823.GK13844@mentat.za.net>

Hi Dorian

On Sat, Dec 01, 2007 at 09:37:42PM +0100, Dorian wrote:
> Hi all,
> 
> Is there any tools about the algorithm of Viterbi using scypi ?.
> Any link on a clear tutorial will be very helpful.

You can find fast dynamic programming / shortest path search code
here:

  http://mentat.za.net/source/shortest_path.tar.bz2

Compile using

  python setup.py build_ext -i

and run the tests to make sure it works.

Regards
St?fan


From arthur00 at gmail.com  Mon Dec  3 11:02:26 2007
From: arthur00 at gmail.com (Arthur Valadares)
Date: Mon, 3 Dec 2007 13:02:26 -0300
Subject: [SciPy-user] Using optimize.fmin_cobyla
Message-ID: <a4447cb30712030802lfbd9b40tc9b4e248c0816482@mail.gmail.com>

Hi, I have the following problem I'm trying to solve:

I have a calculation where, given some parameters, I calculate a vector 'A'
of values and based on this vector, I calculate a final vector 'B' which is
the return of my calculation.
What I want to do is maximize one specific value on 'A' while keeping a
constraint (let C be a constant, the constraint is: C - A[x] = B[x]).

I'm trying to use fmin_cobyla, but I'm a little lost on how to use it to
solve this problem.. Does anyone have any example on using cobyla  or
especifically what I should do? I have tried using it for quite a while, but
I'm just not understanding it right.

On Excel you can just put on Solver to set target A[x], by changing A[x]
with constraint C - A[x] = B[x].

Any tips?

This is my first post, so I wanted to thank the scipy community for all the
great work.

-- 
Arthur Valadares
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071203/70dcf62b/attachment.html>

From aisaac at american.edu  Mon Dec  3 11:43:56 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Mon, 3 Dec 2007 11:43:56 -0500
Subject: [SciPy-user]
	=?utf-8?q?=5Boptimization=5D_BSD-licensed_constraine?=
	=?utf-8?q?d_solver_for=09NLP/NSP?=
In-Reply-To: <4753D704.1090702@scipy.org>
References: <4753D704.1090702@scipy.org>
Message-ID: <Mahogany-0.67.0-1436-20071203-114356.00@american.edu>

On Mon, 03 Dec 2007, dmitrey apparently wrote:
> http://openopt.blogspot.com/2007/12/constraints-handling-for-nlpnsp-solver.html 

It is great that you are succeeding to add constrained 
problems to ralg!

Cheers,
Alan Isaac





From odonnems at yahoo.com  Mon Dec  3 14:06:34 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Mon, 3 Dec 2007 11:06:34 -0800 (PST)
Subject: [SciPy-user] weave and compiler install help
Message-ID: <916245.2008.qm@web58004.mail.re3.yahoo.com>

Thank you for your response Fernando. I am a new python user so this is likely out of my league. I am not sure why the code below is commented out, but I am actually using the blitz. Here is a very simple test that I cannot get to work:

    a = 1
    weave.inline('printf("%d\\n",a);',['a'], verbose=2, type_converters=converters.blitz)


I just added a ticket, as you suggested, under scipy.weave component which will hopefully shed some light. I have tried numerous compilers and versions of scipy/python/enthought. For some reason they all seem to give a similar error. Ticket #550 

I have attached a script denoting all the tests that I have done to get weave.inline to work (this is a .zip file of a py script). The problem for me seems to be that the compiler is creating the source file but does not create the object file. Therefore, the script written by exec_command.py crashes when trying to read the output from this script. I am not sure if the problem lies with the content of the script file or with the compiler. I ran the weave.test() and receive no errors, but I am not sure if it is testing what it really needs to be testing.

Anyhow, thank you for the time and effort and maybe I can figure this out with some additional effort.

Michael O'Donnell



----- Original Message ----
From: Fernando Perez <fperez.net at gmail.com>
To: SciPy Users List <scipy-user at scipy.org>
Sent: Sunday, December 2, 2007 11:09:32 PM
Subject: Re: [SciPy-user] weave and compiler install help


On Nov 19, 2007 9:06 AM, Michael ODonnell <odonnems at yahoo.com> wrote:
>
> 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.

> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 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)

There may be windows-specific problems (I'm on linux), but your code
as above simply can't work because you're assuming blitz behavior
(parens on m(i,j) calls) with the blitz call commented out.

On Linux, this version:

##########################################################
import numpy
from scipy import  weave
from scipy.weave import converters


def prod(m, v):
    #C++ version
    nrows, ncolumns = m.shape
    assert v.ndim==1 and ncolumns==v.shape[0],"Shape mismatch in prod"

    res = numpy.zeros(nrows, 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'],
 verbose=2,
    type_converters=converters.blitz)
    return res

m=numpy.random.rand(3,4)
v = numpy.random.rand(4)
mv = prod(m,v)
mvd = numpy.dot(m,v)
print 'Err:',numpy.linalg.norm(mv-mvd)
#######  EOF  ###############################################

produces errors within floating-point tolerances, as expected:

In [41]: run weaveex.py
Err: 1.11022302463e-16

In [42]: run weaveex.py
Err: 5.55111512313e-17

In [43]: run weaveex.py
Err: 2.77555756156e-17


As I said in my other message, it's still worth filing a ticket on
this with the exact input/output pair that produced the error.  If
it's a core weave bug we'll find it in any platform, and even if it's
platform specific it will ensure that it doesn't get lost in the mists
of mailing lists.

Cheers,

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






      ____________________________________________________________________________________
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/20071203/b7e3dccb/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_weave.zip
Type: application/zip
Size: 3694 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071203/b7e3dccb/attachment.zip>

From wizzard028wise at gmail.com  Mon Dec  3 14:26:52 2007
From: wizzard028wise at gmail.com (Dorian)
Date: Mon, 3 Dec 2007 20:26:52 +0100
Subject: [SciPy-user] Algorithm of Viterbi
In-Reply-To: <20071203152823.GK13844@mentat.za.net>
References: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
	<20071203152823.GK13844@mentat.za.net>
Message-ID: <674a602a0712031126x14f00f61icbc2fb89d110fe63@mail.gmail.com>

Hi Stefan,
I downloaded the files but I don't understand how to compile and for which
example this algorithm works.
Could you please give me a clear explanation step by step how it works ?
I'm using python on Windows .

Thanks in advance

Best regards,

Dorian

PS: I was at  Stellenbosch last year, now I'm in Paris ...

On 03/12/2007, Stefan van der Walt <stefan at sun.ac.za> wrote:
>
> Hi Dorian
>
> On Sat, Dec 01, 2007 at 09:37:42PM +0100, Dorian wrote:
> > Hi all,
> >
> > Is there any tools about the algorithm of Viterbi using scypi ?.
> > Any link on a clear tutorial will be very helpful.
>
> You can find fast dynamic programming / shortest path search code
> here:
>
>   http://mentat.za.net/source/shortest_path.tar.bz2
>
> Compile using
>
>   python setup.py build_ext -i
>
> and run the tests to make sure it works.
>
> Regards
> St?fan
> _______________________________________________
> 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/20071203/1357c64c/attachment.html>

From lorenzo.isella at gmail.com  Mon Dec  3 15:43:21 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Mon, 03 Dec 2007 21:43:21 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <mailman.9.1196618414.1430.scipy-user@scipy.org>
References: <mailman.9.1196618414.1430.scipy-user@scipy.org>
Message-ID: <47546A69.4010700@gmail.com>

Hello,
> Sure. Tell us a bit more about your usecase. Do you want to produce
> figures for print/publication, do you want to interact with your data, or
> do you want simply to do some 3D plotting in a program of yours. You can
> do all this with Mayavi, but the solutions will depend a bit of your usecase.
>
>   
Well, seen that you ask, I often run thermo-hydraulic simulations.
I usually end up having a set of numerical values to plot on a 
non-trivial domain rather than an analytical function.
This can sometimes be done using a contour plot in matplotlib, but I am 
also looking for alternatives.
For instance, let us say that I want to plot some scalar (say a 
component of the velocity) on the cross section of a tube (I simply 
generate random numbers here; in reality I would read it from a file).
With pylab, this is what I would do:

#! /usr/bin/env python
from scipy import *
import pylab
import numpy


nt=20
nr=20




r=linspace(0.,10.,nr)


theta=linspace(0.,2.*pi,nt)
#print "theta is ", theta
sin_t=sin(theta)
cos_t=cos(theta)
rsin_t=r[newaxis,:]*sin_t[:,newaxis]
rcos_t=r[newaxis,:]*cos_t[:,newaxis]
rsin_t=ravel(rsin_t)
rcos_t=ravel(rcos_t)

rsin_t.shape=(nt,nr)
rcos_t.shape=(nt,nr)

vel_section=numpy.random.normal(0.,5.,(nr*nt))

vel_section=reshape(vel_section,(nt,nr))
print 'OK up to here'

#pylab.colorbar()
#pylab.clf()
pylab.figure()
X = rsin_t.transpose()
Y = rcos_t.transpose()
Z = vel_section.transpose()
velmin = vel_section.min()
velmax = vel_section.max()
print velmin, velmax
levels = arange(velmin, velmax+0.01, 0.1)
pylab.contourf(X, Y, Z, levels, cmap=pylab.cm.jet)
pylab.colorbar()

#pylab.show()

pylab.savefig("velocity_on_section_DNS")
#pylab.hold(False)

Can I do something similar with mayavi? And how?


> /usr/lib/python2.4/site-packages/scipy/misc/__init__.py:25: 
> > DeprecationWarning: ScipyTest is now called NumpyTest; please update 
> > your code
>> My main interest is point (1), i.e. to be sure that my python
>> installation is still OK (and know what to do in case it is not). 
>>     
>
> I think it is OK. What happened is that scipy got updated, and these are
> just harmless, albeit annoying, warnings.
>
>   
Here I am a bit puzzled. True, SciPy 0.6.x has recently been released, 
but on my box (Debian testing) I am still running 0.5.2 and I do not 
recall it having been updated recently...or have I missed it?

Many thanks and good luck with the defense of your PhD thesis.

Lorenzo


From odonnems at yahoo.com  Mon Dec  3 16:30:54 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Mon, 3 Dec 2007 13:30:54 -0800 (PST)
Subject: [SciPy-user] weave and compiler install help
Message-ID: <833211.36710.qm@web58002.mail.re3.yahoo.com>

This may help give someone greater insight into my problem. If I run the command created by python:

cl.exe /c /nologo /Ox /MD /W3 /GX /DNDEBUG -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 /Tpc:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860521.cpp /Foc:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860521.obj /Zm1000

at the cmd prompt I get the following error:
c:\Python24\include\pyconfig.h(30) : fatal error C1083: Cannot open include file: 'io.h': No such file or directory

This command was generated by weave.inline, but I copied it into the cmd window.

Does anyone have any ideas what I should/can do?

Thank you,
michael


----- Original Message ----
From: Fernando Perez <fperez.net at gmail.com>
To: SciPy Users List <scipy-user at scipy.org>
Sent: Sunday, December 2, 2007 11:09:32 PM
Subject: Re: [SciPy-user] weave and compiler install help


On Nov 19, 2007 9:06 AM, Michael ODonnell <odonnems at yahoo.com> wrote:
>
> 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.

> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 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)

There may be windows-specific problems (I'm on linux), but your code
as above simply can't work because you're assuming blitz behavior
(parens on m(i,j) calls) with the blitz call commented out.

On Linux, this version:

##########################################################
import numpy
from scipy import  weave
from scipy.weave import converters


def prod(m, v):
    #C++ version
    nrows, ncolumns = m.shape
    assert v.ndim==1 and ncolumns==v.shape[0],"Shape mismatch in prod"

    res = numpy.zeros(nrows, 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'],
 verbose=2,
    type_converters=converters.blitz)
    return res

m=numpy.random.rand(3,4)
v = numpy.random.rand(4)
mv = prod(m,v)
mvd = numpy.dot(m,v)
print 'Err:',numpy.linalg.norm(mv-mvd)
#######  EOF  ###############################################

produces errors within floating-point tolerances, as expected:

In [41]: run weaveex.py
Err: 1.11022302463e-16

In [42]: run weaveex.py
Err: 5.55111512313e-17

In [43]: run weaveex.py
Err: 2.77555756156e-17


As I said in my other message, it's still worth filing a ticket on
this with the exact input/output pair that produced the error.  If
it's a core weave bug we'll find it in any platform, and even if it's
platform specific it will ensure that it doesn't get lost in the mists
of mailing lists.

Cheers,

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






      ____________________________________________________________________________________
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/20071203/af2ed32a/attachment.html>

From discerptor at gmail.com  Mon Dec  3 17:22:02 2007
From: discerptor at gmail.com (Joshua Lippai)
Date: Mon, 3 Dec 2007 14:22:02 -0800
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: <9911419a0712030203r7d631be8qdd67e2f3acef5038@mail.gmail.com>
References: <9911419a0711291017r22d48580q2c1652855d253c86@mail.gmail.com>
	<474F6E3B.1020209@ar.media.kyoto-u.ac.jp>
	<9911419a0712030203r7d631be8qdd67e2f3acef5038@mail.gmail.com>
Message-ID: <9911419a0712031422y4b96c0d7ke57d5d8177837d5c@mail.gmail.com>

On Dec 3, 2007 2:03 AM, Joshua Lippai <discerptor at gmail.com> wrote:
>
> On Nov 29, 2007 5:58 PM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
> >
> > 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
> >
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
>
> That sounds about right... looking at the build dialog, certain
> commands seem to want to use gnu95 while the others want gfortran.
>
> unifing config_fc, config, build_clib, build_ext, build commands
> --fcompiler options
>   commands have different --fcompiler options: ['gfortran', 'gnu95'],
> using first in list as default
>
> I tried setting the --fcompiler option for all of the commands
> mentioned there to gfortran (at least it changed the default from
> gnu95 to gfortran like I wanted), but it still gives that message.
> Anything else I can do to make this confusion for the compiler go
> away?
>
> Josh
>

Here's the full dialog for build and install, btw (and it seems to not
give me that message anymore, but the error still shows up in
scipy.test(1,10)):

Computer:~/python_sources/scipy Josh$ CC=/usr/bin/gcc CXX=/usr/bin/g++
python setup.py build_src config_fc --fcompiler=gfortran config
--fcompiler=gfortran build_clib --fcompiler=gfortran build_ext
--fcompiler=gfortran build --fcompiler=gfortran
mkl_info:
  libraries mkl,vml,guide not found in
/Library/Frameworks/Python.framework/Versions/2.5/lib
  libraries mkl,vml,guide not found in /usr/local/lib
  libraries mkl,vml,guide not found in /usr/lib
  NOT AVAILABLE

fftw3_info:
  libraries fftw3 not found in
/Library/Frameworks/Python.framework/Versions/2.5/lib
  FOUND:
    libraries = ['fftw3']
    library_dirs = ['/usr/local/lib']
    define_macros = [('SCIPY_FFTW3_H', None)]
    include_dirs = ['/usr/local/include']

djbfft_info:
  NOT AVAILABLE

blas_opt_info:
  FOUND:
    extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
    define_macros = [('NO_ATLAS_INFO', 3)]
    extra_compile_args = ['-faltivec',
'-I/System/Library/Frameworks/vecLib.framework/Headers']

lapack_opt_info:
  FOUND:
    extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
    define_macros = [('NO_ATLAS_INFO', 3)]
    extra_compile_args = ['-faltivec']

non-existing path in 'scipy/linsolve': 'tests'
umfpack_info:
  libraries umfpack not found in
/Library/Frameworks/Python.framework/Versions/2.5/lib
  libraries umfpack not found in /usr/local/lib
  libraries umfpack not found in /usr/lib
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/numpy/distutils/system_info.py:414:
UserWarning:
    UMFPACK sparse solver (http://www.cise.ufl.edu/research/sparse/umfpack/)
    not found. Directories to search for the libraries can be specified in the
    numpy/distutils/site.cfg file (section [umfpack]) or by setting
    the UMFPACK environment variable.
  warnings.warn(self.notfounderror.__doc__)
  NOT AVAILABLE

/Users/Josh/python_sources/scipy/scipy/__init__.py:18: UserWarning:
Module scipy was already imported from
/Users/Josh/python_sources/scipy/scipy/__init__.pyc, but
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages
is being added to sys.path
  import pkg_resources as _pr # activate namespace packages
(manipulates __path__)
running build_src
building py_modules sources
building library "dfftpack" sources
building library "linpack_lite" sources
building library "mach" sources
building library "quadpack" sources
building library "odepack" sources
building library "fitpack" sources
building library "superlu_src" sources
building library "odrpack" sources
building library "minpack" sources
building library "rootfind" sources
building library "c_misc" sources
building library "cephes" sources
building library "mach" sources
building library "toms" sources
building library "amos" sources
building library "cdf" sources
building library "specfun" sources
building library "statlib" sources
building extension "scipy.cluster._vq" sources
building extension "scipy.fftpack._fftpack" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.fftpack.convolve" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.integrate._quadpack" sources
building extension "scipy.integrate._odepack" sources
building extension "scipy.integrate.vode" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.interpolate._fitpack" sources
building extension "scipy.interpolate.dfitpack" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/interpolate/dfitpack-f2pywrappers.f'
to sources.
building extension "scipy.io.numpyio" sources
building extension "scipy.lib.blas.fblas" sources
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/build/src.macosx-10.4-ppc-2.5/scipy/lib/blas/fblas-f2pywrappers.f'
to sources.
building extension "scipy.lib.blas.cblas" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/lib/blas/cblas.pyf' to sources.
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.flapack" sources
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.clapack" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/lib/lapack/clapack.pyf'
to sources.
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.calc_lwork" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.atlas_version" sources
building extension "scipy.linalg.fblas" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/fblas.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/build/src.macosx-10.4-ppc-2.5/scipy/linalg/fblas-f2pywrappers.f'
to sources.
building extension "scipy.linalg.cblas" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/cblas.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg.flapack" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/flapack.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/build/src.macosx-10.4-ppc-2.5/scipy/linalg/flapack-f2pywrappers.f'
to sources.
building extension "scipy.linalg.clapack" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/clapack.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg._flinalg" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg.calc_lwork" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg.atlas_version" sources
building extension "scipy.linalg._iterative" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linsolve._zsuperlu" sources
building extension "scipy.linsolve._dsuperlu" sources
building extension "scipy.linsolve._csuperlu" sources
building extension "scipy.linsolve._ssuperlu" sources
building extension "scipy.linsolve.umfpack.__umfpack" sources
building extension "scipy.odr.__odrpack" sources
building extension "scipy.optimize._minpack" sources
building extension "scipy.optimize._zeros" sources
building extension "scipy.optimize._lbfgsb" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.optimize.moduleTNC" sources
building extension "scipy.optimize._cobyla" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.optimize.minpack2" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.signal.sigtools" sources
building extension "scipy.signal.spline" sources
building extension "scipy.sparse._sparsetools" sources
building extension "scipy.special._cephes" sources
building extension "scipy.special.specfun" sources
f2py options: ['--no-wrap-functions']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.stats.statlib" sources
f2py options: ['--no-wrap-functions']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.stats.futil" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.stats.mvn" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/stats/mvn-f2pywrappers.f'
to sources.
building extension "scipy.ndimage._nd_image" sources
building extension "scipy.ndimage.segment._segmenter" sources
building extension "scipy.stsci.convolve._correlate" sources
building extension "scipy.stsci.convolve._lineshape" sources
building extension "scipy.stsci.image._combine" sources
building data_files sources
running config_fc
unifing config_fc, config, build_clib, build_ext, build commands
--fcompiler options
running config
running build_clib
customize UnixCCompiler
customize UnixCCompiler using build_clib
customize Gnu95FCompiler
Found executable /usr/local/bin/gfortran
customize Gnu95FCompiler using build_clib
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
library 'mach' defined more than once, overwriting build_info
{'sources': ['scipy/integrate/mach/d1mach.f',
'scipy/integrate/mach/i1mach.f', 'scipy/integrate/mach/r1mach.f',
'scipy/integrate/mach/xerror.f'], 'config_fc': {'noopt':
('scipy/integrate/setup.pyc', 1)}, 'source_languages': ['f77']}...
with
{'sources': ['scipy/special/mach/d1mach.f',
'scipy/special/mach/i1mach.f', 'scipy/special/mach/r1mach.f',
'scipy/special/mach/xerror.f'], 'config_fc': {'noopt':
('scipy/special/setup.pyc', 1)}, 'source_languages': ['f77']}...
extending extension 'scipy.linsolve._zsuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
extending extension 'scipy.linsolve._dsuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
extending extension 'scipy.linsolve._csuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
extending extension 'scipy.linsolve._ssuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
customize UnixCCompiler
customize UnixCCompiler using build_ext
customize Gnu95FCompiler
customize Gnu95FCompiler using build_ext
running build
running config_cc
unifing config_cc, config, build_clib, build_ext, build commands
--compiler options
running build_py
copying scipy/__svn_version__.py -> build/lib.macosx-10.4-ppc-2.5/scipy
copying build/src.macosx-10.4-ppc-2.5/scipy/__config__.py ->
build/lib.macosx-10.4-ppc-2.5/scipy
copying scipy/io/mmio.py -> build/lib.macosx-10.4-ppc-2.5/scipy/io


Computer:~/python_sources/scipy Josh$ sudo python setup.py install
Password:
mkl_info:
  libraries mkl,vml,guide not found in
/Library/Frameworks/Python.framework/Versions/2.5/lib
  libraries mkl,vml,guide not found in /usr/local/lib
  libraries mkl,vml,guide not found in /usr/lib
  NOT AVAILABLE

fftw3_info:
  libraries fftw3 not found in
/Library/Frameworks/Python.framework/Versions/2.5/lib
  FOUND:
    libraries = ['fftw3']
    library_dirs = ['/usr/local/lib']
    define_macros = [('SCIPY_FFTW3_H', None)]
    include_dirs = ['/usr/local/include']

djbfft_info:
  NOT AVAILABLE

blas_opt_info:
  FOUND:
    extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
    define_macros = [('NO_ATLAS_INFO', 3)]
    extra_compile_args = ['-faltivec',
'-I/System/Library/Frameworks/vecLib.framework/Headers']

lapack_opt_info:
  FOUND:
    extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
    define_macros = [('NO_ATLAS_INFO', 3)]
    extra_compile_args = ['-faltivec']

non-existing path in 'scipy/linsolve': 'tests'
umfpack_info:
  libraries umfpack not found in
/Library/Frameworks/Python.framework/Versions/2.5/lib
  libraries umfpack not found in /usr/local/lib
  libraries umfpack not found in /usr/lib
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/numpy/distutils/system_info.py:414:
UserWarning:
    UMFPACK sparse solver (http://www.cise.ufl.edu/research/sparse/umfpack/)
    not found. Directories to search for the libraries can be specified in the
    numpy/distutils/site.cfg file (section [umfpack]) or by setting
    the UMFPACK environment variable.
  warnings.warn(self.notfounderror.__doc__)
  NOT AVAILABLE

/Users/Josh/python_sources/scipy/scipy/__init__.py:18: UserWarning:
Module scipy was already imported from
/Users/Josh/python_sources/scipy/scipy/__init__.pyc, but
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages
is being added to sys.path
  import pkg_resources as _pr # activate namespace packages
(manipulates __path__)
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 py_modules sources
building library "dfftpack" sources
building library "linpack_lite" sources
building library "mach" sources
building library "quadpack" sources
building library "odepack" sources
building library "fitpack" sources
building library "superlu_src" sources
building library "odrpack" sources
building library "minpack" sources
building library "rootfind" sources
building library "c_misc" sources
building library "cephes" sources
building library "mach" sources
building library "toms" sources
building library "amos" sources
building library "cdf" sources
building library "specfun" sources
building library "statlib" sources
building extension "scipy.cluster._vq" sources
building extension "scipy.fftpack._fftpack" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.fftpack.convolve" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.integrate._quadpack" sources
building extension "scipy.integrate._odepack" sources
building extension "scipy.integrate.vode" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.interpolate._fitpack" sources
building extension "scipy.interpolate.dfitpack" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/interpolate/dfitpack-f2pywrappers.f'
to sources.
building extension "scipy.io.numpyio" sources
building extension "scipy.lib.blas.fblas" sources
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/build/src.macosx-10.4-ppc-2.5/scipy/lib/blas/fblas-f2pywrappers.f'
to sources.
building extension "scipy.lib.blas.cblas" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/lib/blas/cblas.pyf' to sources.
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.flapack" sources
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.clapack" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/lib/lapack/clapack.pyf'
to sources.
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.calc_lwork" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.lib.lapack.atlas_version" sources
building extension "scipy.linalg.fblas" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/fblas.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/build/src.macosx-10.4-ppc-2.5/scipy/linalg/fblas-f2pywrappers.f'
to sources.
building extension "scipy.linalg.cblas" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/cblas.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg.flapack" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/flapack.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/build/src.macosx-10.4-ppc-2.5/scipy/linalg/flapack-f2pywrappers.f'
to sources.
building extension "scipy.linalg.clapack" sources
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/linalg/clapack.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg._flinalg" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg.calc_lwork" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linalg.atlas_version" sources
building extension "scipy.linalg._iterative" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.linsolve._zsuperlu" sources
building extension "scipy.linsolve._dsuperlu" sources
building extension "scipy.linsolve._csuperlu" sources
building extension "scipy.linsolve._ssuperlu" sources
building extension "scipy.linsolve.umfpack.__umfpack" sources
building extension "scipy.odr.__odrpack" sources
building extension "scipy.optimize._minpack" sources
building extension "scipy.optimize._zeros" sources
building extension "scipy.optimize._lbfgsb" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.optimize.moduleTNC" sources
building extension "scipy.optimize._cobyla" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.optimize.minpack2" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.signal.sigtools" sources
building extension "scipy.signal.spline" sources
building extension "scipy.sparse._sparsetools" sources
building extension "scipy.special._cephes" sources
building extension "scipy.special.specfun" sources
f2py options: ['--no-wrap-functions']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.stats.statlib" sources
f2py options: ['--no-wrap-functions']
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.stats.futil" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
building extension "scipy.stats.mvn" sources
f2py options: []
  adding 'build/src.macosx-10.4-ppc-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.4-ppc-2.5' to include_dirs.
  adding 'build/src.macosx-10.4-ppc-2.5/scipy/stats/mvn-f2pywrappers.f'
to sources.
building extension "scipy.ndimage._nd_image" sources
building extension "scipy.ndimage.segment._segmenter" sources
building extension "scipy.stsci.convolve._correlate" sources
building extension "scipy.stsci.convolve._lineshape" sources
building extension "scipy.stsci.image._combine" sources
building data_files sources
running build_py
copying scipy/__svn_version__.py -> build/lib.macosx-10.4-ppc-2.5/scipy
copying build/src.macosx-10.4-ppc-2.5/scipy/__config__.py ->
build/lib.macosx-10.4-ppc-2.5/scipy
running build_clib
customize UnixCCompiler
customize UnixCCompiler using build_clib
customize NAGFCompiler
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
Could not locate executable g77
customize Gnu95FCompiler
Found executable /usr/local/bin/gfortran
customize Gnu95FCompiler
customize Gnu95FCompiler using build_clib
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
library 'mach' defined more than once, overwriting build_info
{'sources': ['scipy/integrate/mach/d1mach.f',
'scipy/integrate/mach/i1mach.f', 'scipy/integrate/mach/r1mach.f',
'scipy/integrate/mach/xerror.f'], 'config_fc': {'noopt':
('scipy/integrate/setup.pyc', 1)}, 'source_languages': ['f77']}...
with
{'sources': ['scipy/special/mach/d1mach.f',
'scipy/special/mach/i1mach.f', 'scipy/special/mach/r1mach.f',
'scipy/special/mach/xerror.f'], 'config_fc': {'noopt':
('scipy/special/setup.pyc', 1)}, 'source_languages': ['f77']}...
extending extension 'scipy.linsolve._zsuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
extending extension 'scipy.linsolve._dsuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
extending extension 'scipy.linsolve._csuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
extending extension 'scipy.linsolve._ssuperlu' defined_macros with
[('USE_VENDOR_BLAS', 1)]
customize UnixCCompiler
customize UnixCCompiler using build_ext
customize NAGFCompiler
customize AbsoftFCompiler
customize IBMFCompiler
customize IntelFCompiler
customize GnuFCompiler
customize Gnu95FCompiler
customize Gnu95FCompiler
customize Gnu95FCompiler using build_ext
running install_lib
copying build/lib.macosx-10.4-ppc-2.5/scipy/__config__.py ->
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy
copying build/lib.macosx-10.4-ppc-2.5/scipy/__svn_version__.py ->
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy
copying build/lib.macosx-10.4-ppc-2.5/scipy/io/mmio.py ->
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/io
byte-compiling /Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/__config__.py
to __config__.pyc
byte-compiling /Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/__svn_version__.py
to __svn_version__.pyc
byte-compiling /Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/io/mmio.py
to mmio.pyc
running install_data
copying scipy/io/tests/test_mmio.py ->
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy/io/tests/
running install_egg_info
Writing /Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/scipy-0.7.0.dev3612-py2.5.egg-info


From giulio.bottazzi at gmail.com  Mon Dec  3 17:32:07 2007
From: giulio.bottazzi at gmail.com (Giulio Bottazzi)
Date: Mon, 3 Dec 2007 23:32:07 +0100
Subject: [SciPy-user] strange interaction between scipy.stats and
	scipy.optimize.minpack
Message-ID: <9fda5e550712031432w6da16666i27dd82b1c94624d2@mail.gmail.com>

Hi all,
I don't know if it has been already reported but I've noticed a
strange behavior. Consider the following script

#START of fit.py ----------------------

from numpy import array
import scipy.stats
from scipy.optimize.minpack import leastsq

empfreq=array([.31,.57,.12])

def thempdiff(p):
    thfreq   = array([(1-p[0])*(1-p[0]), 2*(1-p[0])*p[0],p[0]*p[0] ])
    return thfreq - empfreq

print leastsq(thempdiff,x0=[.5], full_output=1)

#END of fit.py -------------------------

now running it I obtain a seg fault

> python fit.py
zsh: segmentation fault  python fit.py

If instead

1) I remove the line "import scipy.stats" or

2) I remove the option "full_output=1" from leastsq

the script runs smoothly. I don't understand what's going on. Is
somebody else able to reproduce it? Any help?

Giulio.

-- 
Giulio Bottazzi <giulio.bottazzi at gmail.com>
http://giulio.bottazzi.googlepages.com
PGP Key ID:BAB0A33F


From fperez.net at gmail.com  Mon Dec  3 17:54:21 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Mon, 3 Dec 2007 15:54:21 -0700
Subject: [SciPy-user] weave and compiler install help
In-Reply-To: <833211.36710.qm@web58002.mail.re3.yahoo.com>
References: <833211.36710.qm@web58002.mail.re3.yahoo.com>
Message-ID: <db6b5ecc0712031454w20524caag96cfff604d1ebd6@mail.gmail.com>

On Dec 3, 2007 2:30 PM, Michael ODonnell <odonnems at yahoo.com> wrote:
>
> This may help give someone greater insight into my problem. If I run the
> command created by python:
>
> cl.exe /c /nologo /Ox /MD /W3 /GX /DNDEBUG
> -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
> /Tpc:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860521.cpp
> /Foc:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860521.obj
> /Zm1000
>
> at the cmd prompt I get the following error:
> c:\Python24\include\pyconfig.h(30) : fatal error C1083: Cannot open include
> file: 'io.h': No such file or directory
>
> This command was generated by weave.inline, but I copied it into the cmd
> window.
>
> Does anyone have any ideas what I should/can do?

1. Attach the intermediate files (like
c:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860521.cpp)
to the ticket, so either I or someone else can later download them to
debug the problem directly.  Though I'm going to guess this might be a
local configuration problem somewhere, since no one else is reporting
similar issues (and we do have win32 weave users).  Very strange.

2. At least on this list, please use plain text for email. Html makes
reading your messages a huge pain.

Thanks,

f


From odonnems at yahoo.com  Mon Dec  3 18:27:47 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Mon, 3 Dec 2007 15:27:47 -0800 (PST)
Subject: [SciPy-user] weave and compiler install help
Message-ID: <951975.93965.qm@web58007.mail.re3.yahoo.com>

Thanks Fernando for your suggestion.

(Note: Please use the command below and not the command in the previous posting because the filename changed; obviously the paths will need to be altered as well).

I am attaching the .cpp file that was created using weave.inline and the compiler. I mentioned that when I tried to run this at the command prompt, I got an error about not being able to find the io.h file. I did find this file in /Microsoft Visual C++ Toolkit 2003/include and I have this path set in the environment variables so this seems strange to me. 

cl.exe /c /nologo /Ox /MD /W3 /GX /DNDEBUG -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 /Tpc:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.cpp /Foc:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.obj /Zm1000

Thanks for the assistance,
michael

PS I noticed also that /DNDEBUG is not an option for the SDK window compiler. Could this be a problem or am I wrong. I am using this link: http://msdn2.microsoft.com/en-us/library/610ecb4h(VS.80).aspx to look at the options for SDK compiler. I looked at all the versions and did not see this as an option for any of them.





      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071203/654e7506/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sc_552cccf5dbf4f6eadd273cdcbd5860520.zip
Type: application/zip
Size: 3731 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071203/654e7506/attachment.zip>

From odonnems at yahoo.com  Mon Dec  3 18:50:59 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Mon, 3 Dec 2007 15:50:59 -0800 (PST)
Subject: [SciPy-user] weave and compiler install help
Message-ID: <389288.11012.qm@web58005.mail.re3.yahoo.com>

Is there a chance I need to set an environment variable that points to the SDK includes. I noticed that if I copied over C:\Program Files\Microsoft Visual C++ Toolkit 2003\include\io.h and ran the compiler, an error occurred for no float.h files (also found in C:\Program Files\Microsoft Visual C++ Toolkit 2003\include). In the command line / syntax, which I posted, there is no include for this directory, but I would think the compiler would see this. And it still does not explain why mingw or cygwin is not working for me either.


thanks and sorry for all the stupid postings, suggestions from me!
mike




      ____________________________________________________________________________________
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/20071203/5b10f9d0/attachment.html>

From fperez.net at gmail.com  Mon Dec  3 19:13:17 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Mon, 3 Dec 2007 17:13:17 -0700
Subject: [SciPy-user] weave and compiler install help
In-Reply-To: <389288.11012.qm@web58005.mail.re3.yahoo.com>
References: <389288.11012.qm@web58005.mail.re3.yahoo.com>
Message-ID: <db6b5ecc0712031613o5668187dwff0bbb8228ea6446@mail.gmail.com>

On Dec 3, 2007 4:50 PM, Michael ODonnell <odonnems at yahoo.com> wrote:
>
> Is there a chance I need to set an environment variable that points to the
> SDK includes. I noticed that if I copied over C:\Program Files\Microsoft
> Visual C++ Toolkit 2003\include\io.h and ran the compiler, an error occurred
> for no float.h files (also found in C:\Program Files\Microsoft Visual C++
> Toolkit 2003\include). In the command line / syntax, which I posted, there
> is no include for this directory, but I would think the compiler would see
> this. And it still does not explain why mingw or cygwin is not working for
> me either.
>
>
> thanks and sorry for all the stupid postings, suggestions from me!

If your compiler isn't finding float.h, then it really sounds like a
configuration problem.  I'm afraid on that front I don't know how to
help, since I've never used a compiler under Windows, but perhaps
someone else may be able to provide you with some guidance.

Cheers,

f


From matthieu.brucher at gmail.com  Tue Dec  4 01:42:05 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Tue, 4 Dec 2007 07:42:05 +0100
Subject: [SciPy-user] weave and compiler install help
In-Reply-To: <db6b5ecc0712031613o5668187dwff0bbb8228ea6446@mail.gmail.com>
References: <389288.11012.qm@web58005.mail.re3.yahoo.com>
	<db6b5ecc0712031613o5668187dwff0bbb8228ea6446@mail.gmail.com>
Message-ID: <e76aa17f0712032242o24b4011drf9736efc892b5148@mail.gmail.com>

Last time I tried to used another compiler than GCC with weave, it didn't
work because the Blitz header are not included for another compiler in the
repository. So even if the compiler is correctly installed, you won't be
able to use weave because of this.

Matthieu

2007/12/4, Fernando Perez <fperez.net at gmail.com>:
>
> On Dec 3, 2007 4:50 PM, Michael ODonnell <odonnems at yahoo.com> wrote:
> >
> > Is there a chance I need to set an environment variable that points to
> the
> > SDK includes. I noticed that if I copied over C:\Program Files\Microsoft
> > Visual C++ Toolkit 2003\include\io.h and ran the compiler, an error
> occurred
> > for no float.h files (also found in C:\Program Files\Microsoft Visual
> C++
> > Toolkit 2003\include). In the command line / syntax, which I posted,
> there
> > is no include for this directory, but I would think the compiler would
> see
> > this. And it still does not explain why mingw or cygwin is not working
> for
> > me either.
> >
> >
> > thanks and sorry for all the stupid postings, suggestions from me!
>
> If your compiler isn't finding float.h, then it really sounds like a
> configuration problem.  I'm afraid on that front I don't know how to
> help, since I've never used a compiler under Windows, but perhaps
> someone else may be able to provide you with some guidance.
>
> Cheers,
>
> f
> _______________________________________________
> 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/20071204/d442942d/attachment.html>

From nwagner at iam.uni-stuttgart.de  Tue Dec  4 02:15:12 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Tue, 04 Dec 2007 08:15:12 +0100
Subject: [SciPy-user] strange interaction between scipy.stats
	and	scipy.optimize.minpack
In-Reply-To: <9fda5e550712031432w6da16666i27dd82b1c94624d2@mail.gmail.com>
References: <9fda5e550712031432w6da16666i27dd82b1c94624d2@mail.gmail.com>
Message-ID: <web-92854952@uni-stuttgart.de>

On Mon, 3 Dec 2007 23:32:07 +0100
  "Giulio Bottazzi" <giulio.bottazzi at gmail.com> wrote:
> Hi all,
> I don't know if it has been already reported but I've 
>noticed a
> strange behavior. Consider the following script
> 
> #START of fit.py ----------------------
> 
> from numpy import array
> import scipy.stats
> from scipy.optimize.minpack import leastsq
> 
> empfreq=array([.31,.57,.12])
> 
> def thempdiff(p):
>    thfreq   = array([(1-p[0])*(1-p[0]), 
>2*(1-p[0])*p[0],p[0]*p[0] ])
>    return thfreq - empfreq
> 
> print leastsq(thempdiff,x0=[.5], full_output=1)
> 
> #END of fit.py -------------------------
> 
> now running it I obtain a seg fault
> 
>> python fit.py
> zsh: segmentation fault  python fit.py
> 
> If instead
> 
> 1) I remove the line "import scipy.stats" or
> 
> 2) I remove the option "full_output=1" from leastsq
> 
> the script runs smoothly. I don't understand what's 
>going on. Is
> somebody else able to reproduce it? Any help?
> 
> Giulio.
> 
> -- 
> Giulio Bottazzi <giulio.bottazzi at gmail.com>
> http://giulio.bottazzi.googlepages.com
> PGP Key ID:BAB0A33F
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

  
I cannot reproduce your problem.

>>> numpy.__version__
'1.0.5.dev4528'
>>> scipy.__version__
'0.7.0.dev3610'


python  fit.py
(0.42369157713188349, array([[ 0.46734154]]), {'qtf': 
array([ -4.51744596e-06]), 'nfev': 17, 'fjac': array([[ 
1.46279277, -0.20867347, -0.5792871 ]]), 'fvec': array([ 
0.0221314 , -0.08164595,  0.05951455]), 'ipvt': array([1], 
dtype=int32)}, 'Both actual andpredicted relative 
reductions in the sum of squares\n  are at most 0.000000', 
1)


Nils



From giulio.bottazzi at gmail.com  Tue Dec  4 03:35:01 2007
From: giulio.bottazzi at gmail.com (Giulio Bottazzi)
Date: Tue, 4 Dec 2007 09:35:01 +0100
Subject: [SciPy-user] strange interaction between scipy.stats and
	scipy.optimize.minpack
In-Reply-To: <web-92854952@uni-stuttgart.de>
References: <9fda5e550712031432w6da16666i27dd82b1c94624d2@mail.gmail.com>
	<web-92854952@uni-stuttgart.de>
Message-ID: <9fda5e550712040035t5b1711caj406a5c94a33cf26d@mail.gmail.com>

I recompiled everything from scratch but I still get the seg fault.
I'm using python 2.4.4, numpy 1.0.4 and scipy 0.6 compiled with GCC
ver. 4.1.2 on an amd64 box. I'll wait until the packages get updated
in my distro (gentoo) and I'll check it again. Thank you for the
feedback;.

Giulio.

>
> I cannot reproduce your problem.
>
> >>> numpy.__version__
> '1.0.5.dev4528'
> >>> scipy.__version__
> '0.7.0.dev3610'
>
>
> python  fit.py
> (0.42369157713188349, array([[ 0.46734154]]), {'qtf':
> array([ -4.51744596e-06]), 'nfev': 17, 'fjac': array([[
> 1.46279277, -0.20867347, -0.5792871 ]]), 'fvec': array([
> 0.0221314 , -0.08164595,  0.05951455]), 'ipvt': array([1],
> dtype=int32)}, 'Both actual andpredicted relative
> reductions in the sum of squares\n  are at most 0.000000',
> 1)
>
>
> Nils
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
Giulio Bottazzi <giulio.bottazzi at gmail.com>
http://giulio.bottazzi.googlepages.com
PGP Key ID:BAB0A33F


From giovanni.samaey at cs.kuleuven.be  Tue Dec  4 06:05:14 2007
From: giovanni.samaey at cs.kuleuven.be (Giovanni Samaey)
Date: Tue, 4 Dec 2007 12:05:14 +0100
Subject: [SciPy-user] lil_matrix
In-Reply-To: <A53FBE8F-6299-46FF-9AEB-1D4BF9CB4CE8@cs.kuleuven.be>
References: <A53FBE8F-6299-46FF-9AEB-1D4BF9CB4CE8@cs.kuleuven.be>
Message-ID: <6750B3E0-DCA6-44E8-998E-D6E6DEA64099@cs.kuleuven.be>

Hi all,

I have "solved" the issue by writing

B[:5,:5]=A.tocsr()

This fills up B correctly.  However, this approach appears to be very  
inefficient. (Obviously, I am using *much* larger values than 5.)
It seems that this inefficiency is not due to the conversion but due  
to the filling up of B.  This deduction is made based on the fact that  
adding two lil matrices returns a csr matrix in only a fraction of the  
time the above operation takes.

I am using this approach because I have a very large matrix A in  
sparse format and I want to add a border (and extra row and column).

Any help would be appreciated.

Best,

Giovanni

On 30 Nov 2007, at 18:26, Giovanni Samaey wrote:

> 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
> _______________________________________________
> 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 Dec  4 06:21:39 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 04 Dec 2007 20:21:39 +0900
Subject: [SciPy-user] scipy.scons branch: building numpy and scipy with scons
Message-ID: <47553843.7020709@ar.media.kyoto-u.ac.jp>

Hi,

    I've just reached a first usable scipy.scons branch, so that scipy 
can be built entirely with scons (assuming you build numpy with scons 
too). You can get it from 
http://svn.scipy.org/svn/scipy/branches/scipy.scons. To build it, you 
just need to use numpy.scons branch instead of the trunk, and use 
setupscons.py instead of setup.py. Again, I would be happy to hear about 
failures, success (please report a ticket in this case), etc...

Some of the most interesting things I can think of which work with scons:
    - you can control fortran and C flags from the command line: CFLAGS 
and FFLAGS won't override necessary flags, only optimization flags, so 
you can easily play with warning, optimization flags. For example:

CFLAGS='-W -Wall -Wextra -DDEBUG' FFLAGS='-DDEBUG -W -Wall -Wextra' 
python setupscons build

for debugging will work. No need to care about -fPIC and co, all this is 
handled automatically.
    - dependencies are handled correctly thanks to scons: for example, 
if you change a library (e.g. by using MKL=None to disable mkl), only 
link step will be redone.

platforms known to work
-----------------------

    - linux with gcc/g77 or gcc/gfortran (both atlas and mkl 9 were tested).
    - linux with intel compilers (intel and gnu compilers can also be 
mixed, AFAIK).
    - solaris with sun compilers with sunperf, only tested on indiana.

Notable non working things:
---------------------------
 
    - using netlib BLAS and LAPACK is not supported (only optimized ones 
are available: sunperf, atlas, mkl, and vecLib/Accelerate).
    - parallel build does NOT work (AFAICS, this is because f2py which 
do some things which are not thread-safe, but I have not yet found the 
exact problem).
    - I have not yet implemented umfpack checker, and as such umfpack 
cannot be built yet
    - I have not yet tweaked fortran compiler configurations for 
optimizations except for gnu compilers
    - c++ compilers configurations are not handled either.

cheers,

David


From stefan at sun.ac.za  Tue Dec  4 16:29:25 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Tue, 4 Dec 2007 23:29:25 +0200
Subject: [SciPy-user] Algorithm of Viterbi
In-Reply-To: <674a602a0712031126x14f00f61icbc2fb89d110fe63@mail.gmail.com>
References: <674a602a0712011237q2dfe3ae2oe3334c7eb53a78ae@mail.gmail.com>
	<20071203152823.GK13844@mentat.za.net>
	<674a602a0712031126x14f00f61icbc2fb89d110fe63@mail.gmail.com>
Message-ID: <20071204212925.GF8346@mentat.za.net>

Hi Dorian

On Mon, Dec 03, 2007 at 08:26:52PM +0100, Dorian wrote:
> I downloaded the files but I don't understand how to compile and for which
> example this algorithm works.

Sorry, I don't know how to compile things under Windows.  Maybe some
other people on the list have experience with ctypes on that platform.

> Could you please give me a clear explanation step by step how it works ?
> I'm using python on Windows .

There are two functions, "find" and "remove".  I wrote these to see
how well the seam carving algorithm in

  http://www.shaiavidan.org/papers/imretFinal.pdf

works.  The first finds the shortest path through the image, and the
last removes that path.  It is a trivial matter to modify the
algorithm to look for minimal path cost, instead of minimal
difference cost.

Now that I think about it, there is an approximation to Viterbi that
is even easier to implement -- simply take the maximum position in
each column.  It doesn't give the best path (it is not even guaranteed
to be connected), but often it is a good approximation, and it is easy
to implement.  See for example

  http://www.cs.colostate.edu/~hamiltom/hmm/hmm.txt

I can also suggest the following article, which gives a very good
overview of HMMs and Viterbi:

  http://dx.doi.org/10.1109/5.18626

Cheers
St?fan


From pablo_mitchell at yahoo.com  Tue Dec  4 23:25:17 2007
From: pablo_mitchell at yahoo.com (pablo mitchell)
Date: Tue, 4 Dec 2007 20:25:17 -0800 (PST)
Subject: [SciPy-user] TimeSeries - Questions Re lib/moving_funcs.py
Message-ID: <859102.72857.qm@web30805.mail.mud.yahoo.com>

I have comments/questions regarding this module:

* Why do the comments in the source code indicate that many of the functions are intended only for 1-D arrays (mov_var for example)?  

A very brief look at the source doesn't indicate why this should be.  Also, throwing some quick example code together suing 2-D arrays generates results.  The code does not seem to bomb.  

* The covariance related code is constructed in a way I would not expect.

Specifically, most of the functionality appear as facade functions fed into _mov_var_stddev.  This seems to be a multi-purpose function in the module.  Wouldn't it be more natural to define a general covariance function (ignoring bias or span) with pseudo-code:

cov(x,y) = [x - ave(x)]*[y - ave(y)]/n


The rest of the functions fall out of this like:

var(x) = cov(x,x)
stddev(x) = sqrt[var(x)]
corr(x,y) = cov(x,y)/stddev(x)/stddev(y)


Then add-ons like regression coefficients and zscores become trivial.

* This is a really nice package!  Look forward to seeing it grow.





      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071204/aaf74fe3/attachment.html>

From pgmdevlist at gmail.com  Wed Dec  5 00:07:56 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Wed, 5 Dec 2007 00:07:56 -0500
Subject: [SciPy-user] TimeSeries - Questions Re lib/moving_funcs.py
In-Reply-To: <859102.72857.qm@web30805.mail.mud.yahoo.com>
References: <859102.72857.qm@web30805.mail.mud.yahoo.com>
Message-ID: <200712050007.57108.pgmdevlist@gmail.com>

Hello Pablo,
Yes ! Another user !

> * Why do the comments in the source code indicate that many of the
> functions are intended only for 1-D arrays (mov_var for example)?

Because we're confident it works fine with 1D, but haven't tested things 
thoroughly  enough to be as confident with nD arrays. And we don't really 
want to advertise something we can't deliver.

> A very brief look at the source doesn't indicate why this should be.  Also,
> throwing some quick example code together suing 2-D arrays generates
> results.  The code does not seem to bomb.

Actually, I prefer a code that bombs than one that deceives me into believing 
it works when it does not.

> * The covariance related code is constructed in a way I would not expect.

All the moving functions of this module are wrappers to C functions. We use C 
for performance more than for portability (cf the regular problems with 
installation...). What you suggest sounds like a good idea, I'll kick the 
ball to my co-author, he's the C speaker of the duo.

> * This is a really nice package!  Look forward to seeing it grow.

Thanks a lot for your support ! I wish I could spend more time on the code, 
and less time on writing papers, however... Your suggestions and comments are 
always welcome.


From pablo_mitchell at yahoo.com  Wed Dec  5 00:47:03 2007
From: pablo_mitchell at yahoo.com (pablo mitchell)
Date: Tue, 4 Dec 2007 21:47:03 -0800 (PST)
Subject: [SciPy-user] TimeSeries - Questions Re lib/moving_funcs.py
Message-ID: <883344.95910.qm@web30811.mail.mud.yahoo.com>

The C source code seems able to accommodate my suggestion of making the covariance the central function and the remaining ones the facades.  Just a suggestion.

Another suggestion:  a new class similar to TimeSeriesRecords.  

For my purposes, I would prefer to be able to instantiate a 2-D time-series that can be indexed not just on dates (temporal) but also on variable names (spatial if you will).  For example, a time-series structure like:

            var-1  var-2  ... var-n
    date-1  x-11   x-12   ... x-1n       
    date-2  x-21   x-22   ... x-2n
    ...
    date-m  x-m1   x-m2   ... x-mn

could be sliced 

    time-series[date-i:date-j, :]

as well as

    time-series[:, var-i:var-j]

Thanks again -- P

----- Original Message ----
From: Pierre GM <pgmdevlist at gmail.com>
To: pablo mitchell <pablo.mitchell at cal.berkeley.edu>; SciPy Users List <scipy-user at scipy.org>
Sent: Tuesday, December 4, 2007 9:07:56 PM
Subject: Re: [SciPy-user] TimeSeries - Questions Re lib/moving_funcs.py

Hello Pablo,
Yes ! Another user !

> * Why do the comments in the source code indicate that many of the
> functions are intended only for 1-D arrays (mov_var for example)?

Because we're confident it works fine with 1D, but haven't tested
 things 
thoroughly  enough to be as confident with nD arrays. And we don't
 really 
want to advertise something we can't deliver.

> A very brief look at the source doesn't indicate why this should be.
  Also,
> throwing some quick example code together suing 2-D arrays generates
> results.  The code does not seem to bomb.

Actually, I prefer a code that bombs than one that deceives me into
 believing 
it works when it does not.

> * The covariance related code is constructed in a way I would not
 expect.

All the moving functions of this module are wrappers to C functions. We
 use C 
for performance more than for portability (cf the regular problems with
 
installation...). What you suggest sounds like a good idea, I'll kick
 the 
ball to my co-author, he's the C speaker of the duo.

> * This is a really nice package!  Look forward to seeing it grow.

Thanks a lot for your support ! I wish I could spend more time on the
 code, 
and less time on writing papers, however... Your suggestions and
 comments are 
always welcome.







      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071204/19a31f6d/attachment.html>

From pgmdevlist at gmail.com  Wed Dec  5 01:23:21 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Wed, 5 Dec 2007 01:23:21 -0500
Subject: [SciPy-user] TimeSeries - Questions Re lib/moving_funcs.py
In-Reply-To: <883344.95910.qm@web30811.mail.mud.yahoo.com>
References: <883344.95910.qm@web30811.mail.mud.yahoo.com>
Message-ID: <200712050123.34031.pgmdevlist@gmail.com>

On Wednesday 05 December 2007 00:47:03 pablo mitchell wrote:
> The C source code seems able to accommodate my suggestion of making the
> covariance the central function and the remaining ones the facades.  Just a
> suggestion.

Mmh, you know, I can think about something that would be even better than a 
suggestion ;). Feel free to start implementing something, we could definitely 
use some more hands !

>
> Another suggestion:  a new class similar to TimeSeriesRecords.
>
> For my purposes, I would prefer to be able to instantiate a 2-D time-series
> that can be indexed not just on dates (temporal) but also on variable names
> (spatial if you will).  

Yeah, I'd like to have named columns on an homogeneous 2D array as well. I 
guess that could be doable with an extra dictionary (column position:field 
name), and a tailored __getslice__. But I'd see that more as a numpy 
extension, as an hybrid between a 2D ndarray and a record array. 
However, would it be really efficient ? I have the nagging feeling that the 
current method for slicing on dates can be improved. 



From josegomez at gmx.net  Wed Dec  5 11:18:09 2007
From: josegomez at gmx.net (Jose Luis Gomez Dans)
Date: Wed, 05 Dec 2007 17:18:09 +0100
Subject: [SciPy-user] Efficient "windowing"
Message-ID: <20071205161809.18420@gmx.net>

Hi,
I am trying to apply a simple algorithm to a 2D matrix (an image). What I want to do is, for each pixel, choose the highest (... lowest) value in its 8- or 4-connected neighbours. I have done this using weave.inline, and using a couple of loops, but I was curious if there's some way of doing this using numpy slice syntax? My (allegedly, unelegant) attempts have been versions of the following:

b[1:-1,1:-1] = scipy.array([a[0:-2,1:-1] , a[2:,1:-1] , a[1:-1,0:-2] ,\
      a[1:-1,2:],a[0:-2,0:-2], a[0:-2,2:], a[2:,0:-2], a[2:,2:]],'f').max()

They don't work, because the max() call at the end refers to the whole array, so you are given a constant value array, equal to the max. value of a. Using for loops is very slow when dealing with large arrays.

Thanks!
Jose
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser


From n.martinsen-burrell at wartburg.edu  Wed Dec  5 14:58:57 2007
From: n.martinsen-burrell at wartburg.edu (Neil Martinsen-Burrell)
Date: Wed, 05 Dec 2007 13:58:57 -0600
Subject: [SciPy-user] Efficient "windowing"
In-Reply-To: <mailman.13.1196877606.32296.scipy-user@scipy.org>
Message-ID: <C37C5F21.2B5C%n.martinsen-burrell@wartburg.edu>

> Date: Wed, 05 Dec 2007 17:18:09 +0100
> From: "Jose Luis Gomez Dans" <josegomez at gmx.net>
> Subject: [SciPy-user] Efficient "windowing"

> I am trying to apply a simple algorithm to a 2D matrix (an image). What I want
> to do is, for each pixel, choose the highest (... lowest) value in its 8- or
> 4-connected neighbours. I have done this using weave.inline, and using a
> couple of loops, but I was curious if there's some way of doing this using
> numpy slice syntax? My (allegedly, unelegant) attempts have been versions of
> the following:
> 
> b[1:-1,1:-1] = scipy.array([a[0:-2,1:-1] , a[2:,1:-1] , a[1:-1,0:-2] ,\
>       a[1:-1,2:],a[0:-2,0:-2], a[0:-2,2:], a[2:,0:-2], a[2:,2:]],'f').max()
> 
> They don't work, because the max() call at the end refers to the whole array,
> so you are given a constant value array, equal to the max. value of a. Using
> for loops is very slow when dealing with large arrays.

I use something like this:

import numpy
def _neighbors(i,j):
    return [(i-1,j),(i+1,j),(i,j-1),(i,j+1)]

x=numpy.random.random((5,5))
print x[zip(*_neighbors(2,3))].max()

Where the zip (* ) is the inverse of zip and converts the list of tuples
from _neighbors into a tuple of lists.  This uses numpy's fancy indexing to
special case when the indices are themselves lists.

-Neil

-- 
"Your work is to discover your work and then with all your heart to give
yourself to it." 
  -- Buddha



From robert.kern at gmail.com  Wed Dec  5 15:11:44 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 05 Dec 2007 14:11:44 -0600
Subject: [SciPy-user] Efficient "windowing"
In-Reply-To: <20071205161809.18420@gmx.net>
References: <20071205161809.18420@gmx.net>
Message-ID: <47570600.7080006@gmail.com>

Jose Luis Gomez Dans wrote:
> Hi,
> I am trying to apply a simple algorithm to a 2D matrix (an image). What I want to do is, for each pixel, choose the highest (... lowest) value in its 8- or 4-connected neighbours. I have done this using weave.inline, and using a couple of loops, but I was curious if there's some way of doing this using numpy slice syntax? My (allegedly, unelegant) attempts have been versions of the following:
> 
> b[1:-1,1:-1] = scipy.array([a[0:-2,1:-1] , a[2:,1:-1] , a[1:-1,0:-2] ,\
>       a[1:-1,2:],a[0:-2,0:-2], a[0:-2,2:], a[2:,0:-2], a[2:,2:]],'f').max()
> 
> They don't work, because the max() call at the end refers to the whole array, so you are given a constant value array, equal to the max. value of a. Using for loops is very slow when dealing with large arrays.

.max() takes an `axis` argument. It also takes an `out` argument that will help
you save you from making some large temporaries.

In [1]: from numpy import *

In [2]: n = 256

In [5]: a = random.random((n,n)).astype(float32)

In [7]: b = zeros([n,n], dtype=float32)

In [8]: neighbors = array([a[0:-2,1:-1], a[2:,1:-1], a[1:-1,0:-2], a[1:-1,2:],
a[0:-2,0:-2], a[0:-2,2:], a[2:,0:-2], a[2:,2:]])

In [9]: neighbors.shape
Out[9]: (8, 254, 254)

In [10]: neighbors.max(axis=0, out=b[1:-1,1:-1])
Out[10]:
array([[ 0.94350582,  0.94350582,  0.94350582, ...,  0.98800218,
         0.97231197,  0.61088812],
       [ 0.94350582,  0.82977545,  0.94350582, ...,  0.61088812,
         0.97231197,  0.97231197],
       [ 0.94350582,  0.94350582,  0.94350582, ...,  0.95455241,
         0.95455241,  0.95455241],
       ...,
       [ 0.92812324,  0.92812324,  0.95367759, ...,  0.92305821,
         0.92305821,  0.92305821],
       [ 0.95591497,  0.95591497,  0.95367759, ...,  0.96583498,
         0.85454881,  0.92305821],
       [ 0.92812324,  0.95591497,  0.92812324, ...,  0.96583498,
         0.92305821,  0.92305821]], dtype=float32)

In [11]: b
Out[11]:
array([[ 0.        ,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ],
       [ 0.        ,  0.94350582,  0.94350582, ...,  0.97231197,
         0.61088812,  0.        ],
       [ 0.        ,  0.94350582,  0.82977545, ...,  0.97231197,
         0.97231197,  0.        ],
       ...,
       [ 0.        ,  0.95591497,  0.95591497, ...,  0.85454881,
         0.92305821,  0.        ],
       [ 0.        ,  0.92812324,  0.95591497, ...,  0.92305821,
         0.92305821,  0.        ],
       [ 0.        ,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ]], dtype=float32)


-- 
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 odonnems at yahoo.com  Wed Dec  5 18:02:30 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Wed, 5 Dec 2007 15:02:30 -0800 (PST)
Subject: [SciPy-user] weave inline progress/questions for a new error
Message-ID: <109803.32866.qm@web58014.mail.re3.yahoo.com>

I have recently been posting questions about how to implement scipy.weave.inline on a windows XP machine. I have finally made some progress on how to do this. Sorry for the lack of correct terminology and ignorance, but this is outside of my expertise and I only just learned last week what a .cpp and .obj file was.

First, I did not have my library, include and paths set up exactly right. These corrections allowed me to compile the object files.

Second, there is an error in the numpy/distutils/exec_command.py file with regard to finding the correct python executable file on an NT machine. The code was not right, probably due to differneces in python versions or something. I have altered this code, which has allowed me to run weave, but it then crashes at a later point. The program creates the c++ source file, the object file and the weave object file but appears to crash when converting this to a python dll file. I get the following output using verbose (Please skip below this for additional comments and questions):

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
No module named msvccompiler in numpy.distutils; trying from distutils
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e
running build_ext
running build_src
building extension "sc_552cccf5dbf4f6eadd273cdcbd5860520" sources
No module named msvccompiler in numpy.distutils; trying from distutils
customize MSVCCompiler
customize MSVCCompiler using build_ext
No module named msvccompiler in numpy.distutils; trying from distutils
customize MSVCCompiler
Missing compiler_cxx fix for MSVCCompiler
customize MSVCCompiler using build_ext
building 'sc_552cccf5dbf4f6eadd273cdcbd5860520' extension
compiling C sources
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib\site-packages
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib\site-packages\scipy
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib\site-packages\scipy\weave
creating c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib\site-packages\scipy\weave\scxx
C:\Program Files\Microsoft Visual C++ Toolkit 2003\bin\cl.exe /c /nologo /Ox /MD /W3 /GX /DNDEBUG -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 /Tpc:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.cpp /Foc:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.obj /Zm1000
C:\Program Files\Microsoft Visual C++ Toolkit 2003\bin\cl.exe /c /nologo /Ox /MD /W3 /GX /DNDEBUG -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 /TpC:\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp /Foc:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.obj /Zm1000
C:\Program Files\Microsoft Visual C++ Toolkit 2003\bin\link.exe /DLL /nologo /INCREMENTAL:NO /LIBPATH:C:\Python24\libs /LIBPATH:C:\Python24\PCBuild /EXPORT:initsc_552cccf5dbf4f6eadd273cdcbd5860520 c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.obj c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.obj /OUT:c:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.pyd /IMPLIB:c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.lib
LINK : fatal error LNK1104: cannot open file 'msvcprt.lib'
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 349, in ?
    main()
  File "C:\Documents and Settings\Michael\Application Data\ESRI\ArcToolbox\scripts\test_weave.py", line 227, in main
    weave.inline('printf("%d\\n",a);',['a'], verbose=2, type_converters=converters.blitz)  #, compiler = 'msvc', verbose=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 166, in setup
    raise SystemExit, "error: " + str(msg)
CompileError: error: Command "C:\Program Files\Microsoft Visual C++ Toolkit 2003\bin\link.exe /DLL /nologo /INCREMENTAL:NO /LIBPATH:C:\Python24\libs /LIBPATH:C:\Python24\PCBuild /EXPORT:initsc_552cccf5dbf4f6eadd273cdcbd5860520 c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.obj c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.obj /OUT:c:\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.pyd /IMPLIB:c:\temp\Michael\python24_intermediate\compiler_d41d8cd98f00b204e9800998ecf8427e\Release\temp\Michael\python24_compiled\sc_552cccf5dbf4f6eadd273cdcbd5860520.lib" failed with exit status 1104
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

The error seems to be this: LINK : fatal error LNK1104: cannot open file 'msvcprt.lib

Does anyone know what I may need to do to get this to work. I am writing a document that outlines all the details for what I have done in the event someone else is having these issues. I will post this once I get all this worked out. 

Thank you for those who have been corresponding with me as I muddle through this!!

Michael





      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071205/09679952/attachment.html>

From erik.volz at gmail.com  Wed Dec  5 19:56:07 2007
From: erik.volz at gmail.com (Erik Volz)
Date: Wed, 5 Dec 2007 16:56:07 -0800
Subject: [SciPy-user] += has strange behavior with multidimensional arrays
Message-ID: <a47088770712051656le5010c4m11f5380f2479d2c2@mail.gmail.com>

Sorry if this issue is already well-known.

It seems that += and perhaps *= operators do not have correct behavior
with multidimensional arrays.

Let's say I want to make a simple matrix symmetric across the diagonal.

from pylab import *
y = ones((10,10)) * arange(1,11)
print y
yy = y
print 'This is incorrect:'
y+=transpose(y); print y
print 'But this still works:'
yy = yy+transpose(yy); print yy


Clearly python is modifying the array element by element, not
simultaneously. You might not consider this a bug, but the behavior is
so unintuitive (and unlike matlab) that it would trip up most people.


From robert.kern at gmail.com  Wed Dec  5 20:04:35 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 05 Dec 2007 19:04:35 -0600
Subject: [SciPy-user] += has strange behavior with multidimensional
	arrays
In-Reply-To: <a47088770712051656le5010c4m11f5380f2479d2c2@mail.gmail.com>
References: <a47088770712051656le5010c4m11f5380f2479d2c2@mail.gmail.com>
Message-ID: <47574AA3.4000403@gmail.com>

Erik Volz wrote:
> Sorry if this issue is already well-known.
> 
> It seems that += and perhaps *= operators do not have correct behavior
> with multidimensional arrays.

Well, let's be more clear. It does not have the desired behavior when the RHS
contains a view of the array on the LHS. Otherwise, the behavior is correct.
Multidimensionality has nothing to do with it.

> Let's say I want to make a simple matrix symmetric across the diagonal.
> 
> from pylab import *
> y = ones((10,10)) * arange(1,11)
> print y
> yy = y
> print 'This is incorrect:'
> y+=transpose(y); print y
> print 'But this still works:'
> yy = yy+transpose(yy); print yy
> 
> 
> Clearly python is modifying the array element by element, not
> simultaneously. You might not consider this a bug, but the behavior is
> so unintuitive (and unlike matlab) that it would trip up most people.

It's a straightforward consequence of inplace modification and views. It's not
what you want in this case, but there is no way for Python or numpy to know
this. To get what you want, you have to give up one or the other. Above, you
gave up inplace modification. The other approach is to make a copy:

  y += array(y.T)

-- 
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 prabhu at aero.iitb.ac.in  Thu Dec  6 01:53:42 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Thu, 06 Dec 2007 12:23:42 +0530
Subject: [SciPy-user] surface plotting
In-Reply-To: <47546A69.4010700@gmail.com>
References: <mailman.9.1196618414.1430.scipy-user@scipy.org>
	<47546A69.4010700@gmail.com>
Message-ID: <47579C76.7010108@aero.iitb.ac.in>

Lorenzo Isella wrote:
> Hello,
>> Sure. Tell us a bit more about your usecase. Do you want to produce
>> figures for print/publication, do you want to interact with your data, or
>> do you want simply to do some 3D plotting in a program of yours. You can
>> do all this with Mayavi, but the solutions will depend a bit of your usecase.
>>
>>   
> Well, seen that you ask, I often run thermo-hydraulic simulations.
> I usually end up having a set of numerical values to plot on a 
> non-trivial domain rather than an analytical function.

In such cases you might be better served by mayavi.  The only difficulty 
is that for non-trivial domains you'll have to figure out how best to 
represent your data (structured/unstructured etc.).  In fact if you 
really have velocity data in 3D it is nice to be able to see the flow in 
3D using either velocity vectors or streamlines, both of which mayavi 
supports.

> This can sometimes be done using a contour plot in matplotlib, but I am 
> also looking for alternatives.
> For instance, let us say that I want to plot some scalar (say a 
> component of the velocity) on the cross section of a tube (I simply 
> generate random numbers here; in reality I would read it from a file).
> With pylab, this is what I would do:
[...]
> 
> Can I do something similar with mayavi? And how?

With the script you sent, you could do this:

from enthought.mayavi.tools import mlab
mlab.figure()
s = mlab.surf(X, Y, Z, colormap='jet')

This will give you a 3D surface.  You can't set the contours directly 
but you certainly can modify the returned object or do this from the UI. 
  For example,

s.enable_contours = True
s.contour.filled_contours = True
s.contour.contours = range(-17, 16)

If you have vector data you can use either the flow or quiver mlab 
functions.  More on that in the mlab documentation.

Anyway, attached is a script that does what you need for your example 
with mlab.  It doesn't do anything special that your pylab version 
doesn't but it does give you 3D and opens up other possibilities.

If you have other questions on mlab and datasets for non-trivial domains 
please don't hesitate to ask on the enthought-dev list.

cheers,
prabhu

-------------- next part --------------
A non-text attachment was scrubbed...
Name: tm.py
Type: text/x-python
Size: 974 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071206/7a1882ee/attachment.py>

From gael.varoquaux at normalesup.org  Thu Dec  6 03:48:40 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Thu, 6 Dec 2007 09:48:40 +0100
Subject: [SciPy-user] surface plotting
In-Reply-To: <47579C76.7010108@aero.iitb.ac.in>
References: <mailman.9.1196618414.1430.scipy-user@scipy.org>
	<47546A69.4010700@gmail.com> <47579C76.7010108@aero.iitb.ac.in>
Message-ID: <20071206084840.GB27595@clipper.ens.fr>

On Thu, Dec 06, 2007 at 12:23:42PM +0530, Prabhu Ramachandran wrote:
> With the script you sent, you could do this:

> from enthought.mayavi.tools import mlab
> mlab.figure()
> s = mlab.surf(X, Y, Z, colormap='jet')

> This will give you a 3D surface.  You can't set the contours directly 
> but you certainly can modify the returned object or do this from the UI. 
>  For example,

> s.enable_contours = True
> s.contour.filled_contours = True
> s.contour.contours = range(-17, 16)

See also contour_surf which will give you convenient keyword arguments
for that.

Ga?l


From pieter.cogghe at gmail.com  Thu Dec  6 04:56:29 2007
From: pieter.cogghe at gmail.com (Yuccaplant)
Date: Thu, 6 Dec 2007 01:56:29 -0800 (PST)
Subject: [SciPy-user] slice window out of an array of an size,
	origin and boundary mode 	given
Message-ID: <e02fbe2a-6cd8-4553-bdca-efa269cba8fc@y5g2000hsf.googlegroups.com>

Hi,

I've got an array A[y,x] from witch I want to slice a window size =
(b,a) with the current pixel (y,x) as origin. Just like it's done in
many of the filters of scipy.ndimage.filters. But I can't get to the
code.

Just slicing a window out of the array is not the problem, but for the
boundaries I'm a bit lost in way too much for-loops.

thanks a lot,

Pieter


From lorenzo.isella at gmail.com  Thu Dec  6 09:20:21 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Thu, 6 Dec 2007 15:20:21 +0100
Subject: [SciPy-user] Question about Eggs and easy_install
Message-ID: <a2b3004b0712060620x292f2a55he6e14286ebe5a14d@mail.gmail.com>

Dear All,
This question is more about operating system than SciPy itself.
I recently tried to install mayavi2 on my system (Debian testing). I
think I did a bit of mess, though now scipy fine apart from a few
annoying warnings.
First I tried to use eggs and then the procedure described below:



>One of the easiest ways to get things installed without eggs right now
>is to first install the dependencies -- wxPython 2.6, python-vtk and
>numpy and then grab the tarballs here:

 http://code.enthought.com/downloads/source/ets2.6/
>
>Untar them and build them each

 > cd enthought.traits
 > ./build.sh
 > vi/emacs install.sh
 > ./install.sh


I think that Numpy got updated to a version not yet available in the
standard Debian testing repositories.
My question is: as soon as  newer versions of Numpy and SciPy make it
into Debian testing, will I be able to install them automatically with
the usual apt-get upgrade?
In other words: suppose, by absurd, that tomorrow SciPy 1.0 and Numpy
1.7 are released simultaneously as Debian testing packages; will my
system upgrade to them automatically?
I am simply wondering if I am now "condemned"  to use this
non-standard (for me) procedure for updating SciPy & co.
Of course, I am not blaming the list for the advice I got, just I
realize I may not be experienced enough to delve into this at this
stage.
Many thanks

Lorenzo


From gael.varoquaux at normalesup.org  Thu Dec  6 09:32:09 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Thu, 6 Dec 2007 15:32:09 +0100
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <a2b3004b0712060620x292f2a55he6e14286ebe5a14d@mail.gmail.com>
References: <a2b3004b0712060620x292f2a55he6e14286ebe5a14d@mail.gmail.com>
Message-ID: <20071206143209.GG2075@clipper.ens.fr>

On Thu, Dec 06, 2007 at 03:20:21PM +0100, Lorenzo Isella wrote:
> I think that Numpy got updated to a version not yet available in the
> standard Debian testing repositories.

can you tell us the result of 

import numpy
print numpy.__file__
print numpy.__version__
import scipy
print scipy.__file__
print scipy.__version__

So that we have a better idea of what happened.

You are under Debian testing, right?

Cheers,

Ga?l


From matthieu.brucher at gmail.com  Thu Dec  6 10:35:09 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 6 Dec 2007 16:35:09 +0100
Subject: [SciPy-user] Sparse eigensolver
Message-ID: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>

Hi,

Is there a package for scipy (or a function in scipy directly) that can
compute the eigen values and the eigen vectors of a sparse matrix, like the
eigs function in Malab ?

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/20071206/7e053334/attachment.html>

From lbolla at gmail.com  Thu Dec  6 11:52:47 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Thu, 6 Dec 2007 17:52:47 +0100
Subject: [SciPy-user] Sparse eigensolver
In-Reply-To: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
References: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
Message-ID: <80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>

sure: scipy.sandbox.arpack
it's in the sandbox, so you have to enable it and recompile scipy.
note that there is a patch for arpack.py (google for it).
L.


On 12/6/07, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:
>
> Hi,
>
> Is there a package for scipy (or a function in scipy directly) that can
> compute the eigen values and the eigen vectors of a sparse matrix, like the
> eigs function in Malab ?
>
> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071206/a4c98b18/attachment.html>

From zpincus at stanford.edu  Thu Dec  6 11:58:43 2007
From: zpincus at stanford.edu (Zachary Pincus)
Date: Thu, 6 Dec 2007 11:58:43 -0500
Subject: [SciPy-user] Efficient "windowing"
In-Reply-To: <20071205161809.18420@gmx.net>
References: <20071205161809.18420@gmx.net>
Message-ID: <4061C3AE-F477-4DA2-9D57-C9089270A0F2@stanford.edu>

Hello,

You'll want to look into scipy.ndimage.

Definition:     scipy.ndimage.maximum_filter(input, size=None,  
footprint=None, output=None, mode='reflect', cval=0.0, origin=0)
Docstring:
     Calculates a multi-dimensional maximum filter.

     Either a size or a footprint with the filter must be
     provided. An output array can optionally be provided. The origin
     parameter controls the placement of the filter. The mode parameter
     determines how the array borders are handled, where cval is the
     value when mode is equal to 'constant'.

It handles several different boundary conditions, and you can give it  
a "footprint" mask to just select 4- or 8-connected neighbors for  
each pixel. There's also a generic_filter that will call whatever  
function you want for each pixel+neighborhood.

In the context of maximum filtering, if you want to just ignore the  
boundary condition, just use mode="constant", cval=-numpy.Infinity so  
that the filter ignores pixels outside of the image.

Zach


On Dec 5, 2007, at 11:18 AM, Jose Luis Gomez Dans wrote:

> Hi,
> I am trying to apply a simple algorithm to a 2D matrix (an image).  
> What I want to do is, for each pixel, choose the highest (...  
> lowest) value in its 8- or 4-connected neighbours. I have done this  
> using weave.inline, and using a couple of loops, but I was curious  
> if there's some way of doing this using numpy slice syntax? My  
> (allegedly, unelegant) attempts have been versions of the following:
>
> b[1:-1,1:-1] = scipy.array([a[0:-2,1:-1] , a[2:,1:-1] , a 
> [1:-1,0:-2] ,\
>       a[1:-1,2:],a[0:-2,0:-2], a[0:-2,2:], a[2:,0:-2], a 
> [2:,2:]],'f').max()
>
> They don't work, because the max() call at the end refers to the  
> whole array, so you are given a constant value array, equal to the  
> max. value of a. Using for loops is very slow when dealing with  
> large arrays.
>
> Thanks!
> Jose
> -- 
> Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten
> Browser-Versionen downloaden: http://www.gmx.net/de/go/browser
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From matthieu.brucher at gmail.com  Thu Dec  6 12:01:17 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 6 Dec 2007 18:01:17 +0100
Subject: [SciPy-user] Sparse eigensolver
In-Reply-To: <80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
References: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
	<80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
Message-ID: <e76aa17f0712060901w36ad2e63s7312f59c8a36738e@mail.gmail.com>

Thank you for this. I hope it will be put in the trunk at some point (or in
a scikit) as it would be needed for a manifold learning scikit.

Matthieu

2007/12/6, lorenzo bolla <lbolla at gmail.com>:
>
> sure: scipy.sandbox.arpack
> it's in the sandbox, so you have to enable it and recompile scipy.
> note that there is a patch for arpack.py (google for it).
> L.
>
>
> On 12/6/07, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:
>
> > Hi,
> >
> > Is there a package for scipy (or a function in scipy directly) that can
> > compute the eigen values and the eigen vectors of a sparse matrix, like the
> > eigs function in Malab ?
> >
> > 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
> >
> >
>
> _______________________________________________
> 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/20071206/a4c63028/attachment.html>

From lorenzo.isella at gmail.com  Thu Dec  6 12:49:09 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Thu, 6 Dec 2007 18:49:09 +0100
Subject: [SciPy-user] Question about Eggs and easy_install
Message-ID: <a2b3004b0712060949y6eadd32bha153d89d38d07ac0@mail.gmail.com>

Hello,
Here is the output of those commands:


In [1]: import numpy

In [2]: print numpy.__file__
/usr/lib/python2.4/site-packages/numpy-1.0.4-py2.4-linux-i686.egg/numpy/__init__.pyc

In [3]: print numpy.__version__
1.0.4

In [4]: import scipy
/usr/lib/python2.4/site-packages/scipy/misc/__init__.py:25:
DeprecationWarning: ScipyTest is now called NumpyTest; please update
your code
  test = ScipyTest().test

In [5]: print scipy.__file__
/usr/lib/python2.4/site-packages/scipy/__init__.pyc

In [6]: print scipy.__version__
0.5.2

and yes, I am using Debian testing.
Kind Regards

Lorenzo


>
> Message: 5
> Date: Thu, 6 Dec 2007 15:32:09 +0100
> From: Gael Varoquaux <gael.varoquaux at normalesup.org>
> Subject: Re: [SciPy-user] Question about Eggs and easy_install
> To: SciPy Users List <scipy-user at scipy.org>
> Message-ID: <20071206143209.GG2075 at clipper.ens.fr>
> Content-Type: text/plain; charset=iso-8859-1
>
> On Thu, Dec 06, 2007 at 03:20:21PM +0100, Lorenzo Isella wrote:
> > I think that Numpy got updated to a version not yet available in the
> > standard Debian testing repositories.
>
> can you tell us the result of
>
> import numpy
> print numpy.__file__
> print numpy.__version__
> import scipy
> print scipy.__file__
> print scipy.__version__
>
> So that we have a better idea of what happened.
>
> You are under Debian testing, right?
>
> Cheers,
>
> Ga?l


From lbolla at gmail.com  Thu Dec  6 12:55:04 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Thu, 6 Dec 2007 18:55:04 +0100
Subject: [SciPy-user] Sparse eigensolver
In-Reply-To: <e76aa17f0712060901w36ad2e63s7312f59c8a36738e@mail.gmail.com>
References: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
	<80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
	<e76aa17f0712060901w36ad2e63s7312f59c8a36738e@mail.gmail.com>
Message-ID: <80c99e790712060955i21fa05f5w1e30957ca4ec6193@mail.gmail.com>

I agree...

On Dec 6, 2007 6:01 PM, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:

> Thank you for this. I hope it will be put in the trunk at some point (or
> in a scikit) as it would be needed for a manifold learning scikit.
>
> Matthieu
>
> 2007/12/6, lorenzo bolla < lbolla at gmail.com>:
> >
> > sure: scipy.sandbox.arpack
> > it's in the sandbox, so you have to enable it and recompile scipy.
> > note that there is a patch for arpack.py (google for it).
> > L.
> >
> >
> >  On 12/6/07, Matthieu Brucher < matthieu.brucher at gmail.com> wrote:
> >
> > > Hi,
> > >
> > > Is there a package for scipy (or a function in scipy directly) that
> > > can compute the eigen values and the eigen vectors of a sparse matrix, like
> > > the eigs function in Malab ?
> > >
> > > 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
> > >
> > >
> >
> > _______________________________________________
> > 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
>
> _______________________________________________
> 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/20071206/3d6e4b3d/attachment.html>

From skraelings001 at gmail.com  Thu Dec  6 13:04:55 2007
From: skraelings001 at gmail.com (Reynaldo Baquerizo)
Date: Thu, 06 Dec 2007 13:04:55 -0500
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <mailman.6108.1196960509.2301.scipy-user@scipy.org>
References: <mailman.6108.1196960509.2301.scipy-user@scipy.org>
Message-ID: <475839C7.701@gmail.com>


> I think that Numpy got updated to a version not yet available in the
> standard Debian testing repositories.
> My question is: as soon as  newer versions of Numpy and SciPy make it
> into Debian testing, will I be able to install them automatically with
> the usual apt-get upgrade?
> In other words: suppose, by absurd, that tomorrow SciPy 1.0 and Numpy
> 1.7 are released simultaneously as Debian testing packages; will my
> system upgrade to them automatically?
>   

In general, any package that you build yourself is not updated cause 
'apt' isn't aware of its presence.
So, if you want to update to a recent version you have to do it manually.

Be aware of, if you are keeping too different versions of scipy, numpy, 
etc ..one maintained by 'apt' and the other by you , then only the one 
maintained by 'apt' will be updated.



> I am simply wondering if I am now "condemned"  to use this
> non-standard (for me) procedure for updating SciPy & co.
> Of course, I am not blaming the list for the advice I got, just I
> realize I may not be experienced enough to delve into this at this
> stage.
> Many thanks
>
> Lorenzo
>
>
>   



From prabhu at aero.iitb.ac.in  Thu Dec  6 13:14:48 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Thu, 06 Dec 2007 23:44:48 +0530
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <a2b3004b0712060949y6eadd32bha153d89d38d07ac0@mail.gmail.com>
References: <a2b3004b0712060949y6eadd32bha153d89d38d07ac0@mail.gmail.com>
Message-ID: <47583C18.7050100@aero.iitb.ac.in>

Lorenzo Isella wrote:
> In [1]: import numpy
> 
> In [2]: print numpy.__file__
> /usr/lib/python2.4/site-packages/numpy-1.0.4-py2.4-linux-i686.egg/numpy/__init__.pyc

It looks like your numpy has been upgraded.  One way to undo this is to do:

  easy_install -m numpy

and then:

rm -rf /usr/lib/python2.4/site-packages/numpy-1.0.4-py2.4-linux-i686.egg


This should hopefully revert back to your debian version of numpy.  If 
it doesn't do an apt-get install --reinstall python-numpy to reinstall 
the debian version.

HTH.
prabhu


From prabhu at aero.iitb.ac.in  Thu Dec  6 13:17:36 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Thu, 06 Dec 2007 23:47:36 +0530
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <a2b3004b0712060620x292f2a55he6e14286ebe5a14d@mail.gmail.com>
References: <a2b3004b0712060620x292f2a55he6e14286ebe5a14d@mail.gmail.com>
Message-ID: <47583CC0.9040405@aero.iitb.ac.in>

Lorenzo Isella wrote:
> I think that Numpy got updated to a version not yet available in the
> standard Debian testing repositories.
> My question is: as soon as  newer versions of Numpy and SciPy make it
> into Debian testing, will I be able to install them automatically with
> the usual apt-get upgrade?

If you remove the numpy-1.0.4 egg, then yes you shouldn't have any trouble.

> In other words: suppose, by absurd, that tomorrow SciPy 1.0 and Numpy
> 1.7 are released simultaneously as Debian testing packages; will my
> system upgrade to them automatically?

They should.

> I am simply wondering if I am now "condemned"  to use this
> non-standard (for me) procedure for updating SciPy & co.

No.  You will have trouble when mayavi2 gets into debian though.  When 
that happens you may just want to blow away all the enthought.* eggs 
that you have already installed.

Getting the mayavi2 package into testing may take a little while but a 
few kind folks are working on the packaging currently.

cheers,
prabhu


From matthieu.brucher at gmail.com  Thu Dec  6 14:23:29 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 6 Dec 2007 20:23:29 +0100
Subject: [SciPy-user] Sparse eigensolver
In-Reply-To: <80c99e790712060955i21fa05f5w1e30957ca4ec6193@mail.gmail.com>
References: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
	<80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
	<e76aa17f0712060901w36ad2e63s7312f59c8a36738e@mail.gmail.com>
	<80c99e790712060955i21fa05f5w1e30957ca4ec6193@mail.gmail.com>
Message-ID: <e76aa17f0712061123s69d1231bqce047126dbbe94c7@mail.gmail.com>

Well, if I need it (I can use a dense one for the moment) and if people
agree, I'll make it a scikit ;)

Matthieu

2007/12/6, lorenzo bolla <lbolla at gmail.com>:
>
> I agree...
>
> On Dec 6, 2007 6:01 PM, Matthieu Brucher <matthieu.brucher at gmail.com>
> wrote:
>
> > Thank you for this. I hope it will be put in the trunk at some point (or
> > in a scikit) as it would be needed for a manifold learning scikit.
> >
> > Matthieu
> >
> > 2007/12/6, lorenzo bolla < lbolla at gmail.com>:
> > >
> > > sure: scipy.sandbox.arpack
> > > it's in the sandbox, so you have to enable it and recompile scipy.
> > > note that there is a patch for arpack.py (google for it).
> > > L.
> > >
> > >
> > >  On 12/6/07, Matthieu Brucher < matthieu.brucher at gmail.com> wrote:
> > >
> > > > Hi,
> > > >
> > > > Is there a package for scipy (or a function in scipy directly) that
> > > > can compute the eigen values and the eigen vectors of a sparse matrix, like
> > > > the eigs function in Malab ?
> > > >
> > > > 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
> > > >
> > > >
> > >
> > > _______________________________________________
> > > 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
> >
> > _______________________________________________
> > 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
>
>


-- 
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/20071206/4275863d/attachment.html>

From Alexander.Dietz at astro.cf.ac.uk  Thu Dec  6 15:56:50 2007
From: Alexander.Dietz at astro.cf.ac.uk (Alexander Dietz)
Date: Thu, 6 Dec 2007 20:56:50 +0000
Subject: [SciPy-user] Problems installing scipy 0.6 on FC5
In-Reply-To: <474BE783.2020707@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>
	<474BE783.2020707@ar.media.kyoto-u.ac.jp>
Message-ID: <9cf809a00712061256x5a99159fwe5f933b16af78e7f@mail.gmail.com>

Hi,

I finally found a work-around for my problems installing scipy for my Linux
distribution:
I needed just to comment the line in

/numpy/distutils/fcompiler/gnu.py

in which the term "msse2" is set as an compiler option. Then scipy installs
without problems...

Cheers
  Alex

On Nov 27, 2007 9:46 AM, David Cournapeau <david at ar.media.kyoto-u.ac.jp>
wrote:

> 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
> _______________________________________________
> 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/20071206/6fd3e4cd/attachment.html>

From dueck at math.ucdavis.edu  Thu Dec  6 23:04:45 2007
From: dueck at math.ucdavis.edu (Pierre Dueck)
Date: Thu, 6 Dec 2007 20:04:45 -0800
Subject: [SciPy-user] I'm lost on installation
Message-ID: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>

I'm trying to install Numpy and Scipy.  I'm lost on how to finish this  
- I would appreciate help very much!  I found a problem when running  
the numpy test:

 >>> import numpy
Running from numpy source directory.
 >>> numpy.test(1,10)
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'test'
 >>>

Thanks!
-Pierre

Here is the requested info:

Your OS version
	OS X 10.5 Leopard

gcc --version
	i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5465)

gfortran --version
	GNU Fortran (GCC) 4.2.1

The versions of numpy and scipy that you are trying to install
	Numpy 1.0.4
	Scipy 0.6.0

The full output of $ python setup.py build:

$ python setup.py build
Running from numpy source directory.
non-existing path in 'numpy/distutils': 'site.cfg'
F2PY Version 2_4422
blas_opt_info:
   FOUND:
     extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
     define_macros = [('NO_ATLAS_INFO', 3)]
     extra_compile_args = ['-faltivec', '-I/System/Library/Frameworks/ 
vecLib.framework/Headers']

lapack_opt_info:
   FOUND:
     extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
     define_macros = [('NO_ATLAS_INFO', 3)]
     extra_compile_args = ['-faltivec']

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 py_modules sources
building extension "numpy.core.multiarray" sources
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/config.h' to  
sources.
executing numpy/core/code_generators/generate_array_api.py
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/ 
__multiarray_api.h' to sources.
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/src' to  
include_dirs.
numpy.core - nothing done with h_files = ['build/src.macosx-10.3- 
fat-2.5/numpy/core/src/scalartypes.inc', 'build/src.macosx-10.3- 
fat-2.5/numpy/core/src/arraytypes.inc', 'build/src.macosx-10.3-fat-2.5/ 
numpy/core/config.h', 'build/src.macosx-10.3-fat-2.5/numpy/core/ 
__multiarray_api.h']
building extension "numpy.core.umath" sources
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/config.h' to  
sources.
executing numpy/core/code_generators/generate_ufunc_api.py
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/__ufunc_api.h' to  
sources.
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/src' to  
include_dirs.
numpy.core - nothing done with h_files = ['build/src.macosx-10.3- 
fat-2.5/numpy/core/src/scalartypes.inc', 'build/src.macosx-10.3- 
fat-2.5/numpy/core/src/arraytypes.inc', 'build/src.macosx-10.3-fat-2.5/ 
numpy/core/config.h', 'build/src.macosx-10.3-fat-2.5/numpy/core/ 
__ufunc_api.h']
building extension "numpy.core._sort" sources
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/config.h' to  
sources.
executing numpy/core/code_generators/generate_array_api.py
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/ 
__multiarray_api.h' to sources.
numpy.core - nothing done with h_files = ['build/src.macosx-10.3- 
fat-2.5/numpy/core/config.h', 'build/src.macosx-10.3-fat-2.5/numpy/ 
core/__multiarray_api.h']
building extension "numpy.core.scalarmath" sources
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/config.h' to  
sources.
executing numpy/core/code_generators/generate_array_api.py
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/ 
__multiarray_api.h' to sources.
executing numpy/core/code_generators/generate_ufunc_api.py
   adding 'build/src.macosx-10.3-fat-2.5/numpy/core/__ufunc_api.h' to  
sources.
numpy.core - nothing done with h_files = ['build/src.macosx-10.3- 
fat-2.5/numpy/core/config.h', 'build/src.macosx-10.3-fat-2.5/numpy/ 
core/__multiarray_api.h', 'build/src.macosx-10.3-fat-2.5/numpy/core/ 
__ufunc_api.h']
building extension "numpy.core._dotblas" sources
   adding 'numpy/core/blasdot/_dotblas.c' to sources.
building extension "numpy.lib._compiled_base" sources
building extension "numpy.numarray._capi" sources
building extension "numpy.fft.fftpack_lite" sources
building extension "numpy.linalg.lapack_lite" sources
   adding 'numpy/linalg/lapack_litemodule.c' to sources.
building extension "numpy.random.mtrand" sources
customize NAGFCompiler
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
Could not locate executable g77
customize Gnu95FCompiler
Found executable /usr/local/bin/gfortran
customize Gnu95FCompiler
customize Gnu95FCompiler using config
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 -fPIC -fno-common -dynamic -DNDEBUG -g -O3 -Wall - 
Wstrict-prototypes

compile options: '-Inumpy/core/src -Inumpy/core/include -I/Library/ 
Frameworks/Python.framework/Versions/2.5/include/python2.5 -c'
gcc: _configtest.c
gcc _configtest.o -o _configtest
_configtest
failure.
removing: _configtest.c _configtest.o _configtest
building data_files sources
running build_py
copying build/src.macosx-10.3-fat-2.5/numpy/__config__.py -> build/ 
lib.macosx-10.3-fat-2.5/numpy
copying build/src.macosx-10.3-fat-2.5/numpy/distutils/__config__.py ->  
build/lib.macosx-10.3-fat-2.5/numpy/distutils
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
running build_scripts
   adding 'build/scripts.macosx-10.3-fat-2.5/f2py' to scripts

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071206/083f50a3/attachment.html>

From david at ar.media.kyoto-u.ac.jp  Thu Dec  6 22:58:13 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 07 Dec 2007 12:58:13 +0900
Subject: [SciPy-user] I'm lost on installation
In-Reply-To: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>
References: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>
Message-ID: <4758C4D5.9010104@ar.media.kyoto-u.ac.jp>

Pierre Dueck wrote:
> I'm trying to install Numpy and Scipy.  I'm lost on how to finish this 
> - I would appreciate help very much!  I found a problem when running 
> the numpy test:
>
> >>> import numpy
> Running from numpy source directory.
> >>> numpy.test(1,10)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in <module>
> AttributeError: 'module' object has no attribute 'test'
> >>>
You should not launch python while you are in the source tree. The 
message "running from numpy source directory" may not be really clear, 
but this means numpy has no chance of working for normal usage. Same for 
scipy: once you've finished installing scipy, you should not try to 
launch python in the source directory.

cheers,

David


From robert.kern at gmail.com  Thu Dec  6 23:09:56 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 06 Dec 2007 22:09:56 -0600
Subject: [SciPy-user] I'm lost on installation
In-Reply-To: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>
References: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>
Message-ID: <4758C794.2010405@gmail.com>

Pierre Dueck wrote:
> I'm trying to install Numpy and Scipy.  I'm lost on how to finish this -
> I would appreciate help very much!  I found a problem when running the
> numpy test:
> 
>>>> import numpy
> Running from numpy source directory.

Change to another directory before trying to import numpy. Otherwise, Python
will pick up the unbuilt sources instead of the installed package since Python
searches the current directory before the installation location.

-- 
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 dueck at math.ucdavis.edu  Thu Dec  6 23:14:49 2007
From: dueck at math.ucdavis.edu (Pierre Dueck)
Date: Thu, 6 Dec 2007 20:14:49 -0800
Subject: [SciPy-user] I'm lost on installation
In-Reply-To: <4758C4D5.9010104@ar.media.kyoto-u.ac.jp>
References: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>
	<4758C4D5.9010104@ar.media.kyoto-u.ac.jp>
Message-ID: <360373E8-B8C1-47A5-B4F7-DBD17DA5C361@math.ucdavis.edu>

Thanks!  That was fast!

-Pierre

On Dec 6, 2007, at 7:58 PM, David Cournapeau wrote:

> Pierre Dueck wrote:
>> I'm trying to install Numpy and Scipy.  I'm lost on how to finish  
>> this
>> - I would appreciate help very much!  I found a problem when running
>> the numpy test:
>>
>>>>> import numpy
>> Running from numpy source directory.
>>>>> numpy.test(1,10)
>> Traceback (most recent call last):
>>  File "<stdin>", line 1, in <module>
>> AttributeError: 'module' object has no attribute 'test'
>>>>>
> You should not launch python while you are in the source tree. The
> message "running from numpy source directory" may not be really clear,
> but this means numpy has no chance of working for normal usage. Same  
> for
> scipy: once you've finished installing scipy, you should not try to
> launch python in the source directory.
>
> 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  Thu Dec  6 23:06:54 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 07 Dec 2007 13:06:54 +0900
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <47583C18.7050100@aero.iitb.ac.in>
References: <a2b3004b0712060949y6eadd32bha153d89d38d07ac0@mail.gmail.com>
	<47583C18.7050100@aero.iitb.ac.in>
Message-ID: <4758C6DE.3060205@ar.media.kyoto-u.ac.jp>

Prabhu Ramachandran wrote:
> Lorenzo Isella wrote:
>   
>> In [1]: import numpy
>>
>> In [2]: print numpy.__file__
>> /usr/lib/python2.4/site-packages/numpy-1.0.4-py2.4-linux-i686.egg/numpy/__init__.pyc
>>     
>
> It looks like your numpy has been upgraded.  One way to undo this is to do:
>
>   easy_install -m numpy
>
> and then:
>
> rm -rf /usr/lib/python2.4/site-packages/numpy-1.0.4-py2.4-linux-i686.egg
>
>
> This should hopefully revert back to your debian version of numpy.  If 
> it doesn't do an apt-get install --reinstall python-numpy to reinstall 
> the debian version.
>
>   
The whole problem is coming from the fact that Lorenzo installed things 
in /usr. To keep things simple: never ever do that. For example, if you 
install numpy from debian, and then overwrite with your own numpy, it is 
possible that you won't be able to remove the package (the scripts to 
remove the packages may rely on some behaviour which is different 
between the packaged numpy and your own). dpkg won't overwrite, though, 
because by default, it fails if a package tries to overwrite any 
existing file, but that does not make it a good idea.

The only reliable way to have dpkg-aware numpy and your own is to 
install your own in a different place. Doing otherwise is really likely 
to give you some headache at some points.

cheers,

David


From lev at vpac.org  Fri Dec  7 00:10:36 2007
From: lev at vpac.org (Lev Lafayette)
Date: Fri, 07 Dec 2007 16:10:36 +1100
Subject: [SciPy-user] I'm lost on installation
In-Reply-To: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>
References: <EA2EC449-5778-4453-B600-487B6502143A@math.ucdavis.edu>
Message-ID: <1197004236.4934.72.camel@sys09.in.vpac.org>

On Thu, 2007-12-06 at 20:04 -0800, Pierre Dueck wrote:
> I'm trying to install Numpy and Scipy.  I'm lost on how to finish this
> - I would appreciate help very much!  I found a problem when running
> the numpy test:
> 
> 
> >>> import numpy
> Running from numpy source directory.

There's the answer. Go up a directory and run the test again.

All the best,


Lev
> 



From prabhu at aero.iitb.ac.in  Fri Dec  7 00:35:06 2007
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Fri, 07 Dec 2007 11:05:06 +0530
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <4758C6DE.3060205@ar.media.kyoto-u.ac.jp>
References: <a2b3004b0712060949y6eadd32bha153d89d38d07ac0@mail.gmail.com>	<47583C18.7050100@aero.iitb.ac.in>
	<4758C6DE.3060205@ar.media.kyoto-u.ac.jp>
Message-ID: <4758DB8A.9000002@aero.iitb.ac.in>

David Cournapeau wrote:
> The whole problem is coming from the fact that Lorenzo installed things 
> in /usr. To keep things simple: never ever do that. For example, if you 

On Debian/debian-derivatives the recommended way is to simply do

easy_install --prefix=/usr/local [other args]

and this will work.  On other distros this may not work. FC6 is one such 
beast where this does not work.

If you have a ton of stuff in /usr/local and have a hard time managing 
those, I'd recommend using GNU stow.

cheers,
prabhu



From david at ar.media.kyoto-u.ac.jp  Fri Dec  7 01:07:40 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 07 Dec 2007 15:07:40 +0900
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <4758DB8A.9000002@aero.iitb.ac.in>
References: <a2b3004b0712060949y6eadd32bha153d89d38d07ac0@mail.gmail.com>	<47583C18.7050100@aero.iitb.ac.in>	<4758C6DE.3060205@ar.media.kyoto-u.ac.jp>
	<4758DB8A.9000002@aero.iitb.ac.in>
Message-ID: <4758E32C.3010702@ar.media.kyoto-u.ac.jp>

Prabhu Ramachandran wrote:
> David Cournapeau wrote:
>   
>> The whole problem is coming from the fact that Lorenzo installed things 
>> in /usr. To keep things simple: never ever do that. For example, if you 
>>     
>
> On Debian/debian-derivatives the recommended way is to simply do
>
> easy_install --prefix=/usr/local [other args]
>   
/usr/local is indeed supposed to be the default path for such things 
(that's the default path used by GNU build system, for example), and the 
one where to install things not under the control of your distribution 
(there is /opt, too; but I don't think it matters much which one you use).

David


From fperez.net at gmail.com  Fri Dec  7 02:21:48 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Fri, 7 Dec 2007 00:21:48 -0700
Subject: [SciPy-user] Scipy.org having problems?
Message-ID: <db6b5ecc0712062321j2aaa40fft1663aa14490d22c8@mail.gmail.com>

Hi all,

the main page is working, but access to subpages (such as the
cookbook) seems to be stalling out, and eventually:

OK

The server encountered an internal error or misconfiguration and was
unable to complete your request.

Please contact the server administrator, root at localhost and inform
them of the time the error occurred, and anything you might have done
that may have caused the error.

More information about this error may be available in the server error log.


Any ideas?

Thanks!

f


From robert.kern at gmail.com  Fri Dec  7 02:24:53 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 07 Dec 2007 01:24:53 -0600
Subject: [SciPy-user] Scipy.org having problems?
In-Reply-To: <db6b5ecc0712062321j2aaa40fft1663aa14490d22c8@mail.gmail.com>
References: <db6b5ecc0712062321j2aaa40fft1663aa14490d22c8@mail.gmail.com>
Message-ID: <4758F545.9020503@gmail.com>

Fernando Perez wrote:
> Hi all,
> 
> the main page is working, but access to subpages (such as the
> cookbook) seems to be stalling out, and eventually:
> 
> OK
> 
> The server encountered an internal error or misconfiguration and was
> unable to complete your request.
> 
> Please contact the server administrator, root at localhost and inform
> them of the time the error occurred, and anything you might have done
> that may have caused the error.
> 
> More information about this error may be available in the server error log.
> 
> 
> Any ideas?

It's working fine for me, but I've forwarded your message on to Ryan.

-- 
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 fperez.net at gmail.com  Fri Dec  7 02:38:34 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Fri, 7 Dec 2007 00:38:34 -0700
Subject: [SciPy-user] Scipy.org having problems?
In-Reply-To: <4758F545.9020503@gmail.com>
References: <db6b5ecc0712062321j2aaa40fft1663aa14490d22c8@mail.gmail.com>
	<4758F545.9020503@gmail.com>
Message-ID: <db6b5ecc0712062338g608153f6s789b496709384606@mail.gmail.com>

On Dec 7, 2007 12:24 AM, Robert Kern <robert.kern at gmail.com> wrote:

> It's working fine for me, but I've forwarded your message on to Ryan.

Thanks!

The load average was hovering around 7 a moment ago, with a bunch of
trac.fcgi processes, in case it helps.

M    TIME+  COMMAND
31517 scipy     15   0 50312  19m 4232 S 42.1  1.0   2:08.46 trac.fcgi
23756 scipy     15   0 88872  19m 4232 S 23.2  1.0  18:17.42 trac.fcgi
 1800 ipython   15   0 42316  15m 3952 S 18.9  0.8   0:24.71 trac.fcgi
30365 scipy     15   0 42340  14m 4292 S  7.6  0.7   0:29.47 trac.fcgi
27390 astropy   16   0 62168  22m 3960 S  6.0  1.1  32:02.84 trac.fcgi
24035 scipy     15   0 41784  12m 4144 S  4.6  0.6   0:45.02 trac.fcgi


Cheers,

f


From matthieu.brucher at gmail.com  Fri Dec  7 03:26:23 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 7 Dec 2007 09:26:23 +0100
Subject: [SciPy-user] Scipy.org having problems?
In-Reply-To: <db6b5ecc0712062338g608153f6s789b496709384606@mail.gmail.com>
References: <db6b5ecc0712062321j2aaa40fft1663aa14490d22c8@mail.gmail.com>
	<4758F545.9020503@gmail.com>
	<db6b5ecc0712062338g608153f6s789b496709384606@mail.gmail.com>
Message-ID: <e76aa17f0712070026u71af3ca2mf69208d03abc666e@mail.gmail.com>

One hour ago, the main site did not respond for me.

Matthieu

2007/12/7, Fernando Perez <fperez.net at gmail.com>:
>
> On Dec 7, 2007 12:24 AM, Robert Kern <robert.kern at gmail.com> wrote:
>
> > It's working fine for me, but I've forwarded your message on to Ryan.
>
> Thanks!
>
> The load average was hovering around 7 a moment ago, with a bunch of
> trac.fcgi processes, in case it helps.
>
> M    TIME+  COMMAND
> 31517 scipy     15   0 50312  19m 4232 S 42.1  1.0   2:08.46 trac.fcgi
> 23756 scipy     15   0 88872  19m 4232 S 23.2  1.0  18:17.42 trac.fcgi
> 1800 ipython   15   0 42316  15m 3952 S 18.9  0.8   0:24.71 trac.fcgi
> 30365 scipy     15   0 42340  14m 4292 S  7.6  0.7   0:29.47 trac.fcgi
> 27390 astropy   16   0 62168  22m 3960 S  6.0  1.1  32:02.84 trac.fcgi
> 24035 scipy     15   0 41784  12m 4144 S  4.6  0.6   0:45.02 trac.fcgi
>
>
> Cheers,
>
> f
> _______________________________________________
> 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/20071207/8c632914/attachment.html>

From shuntim.luk at polyu.edu.hk  Fri Dec  7 03:41:33 2007
From: shuntim.luk at polyu.edu.hk (LUK ShunTim)
Date: Fri, 07 Dec 2007 16:41:33 +0800
Subject: [SciPy-user] Question about Eggs and easy_install
In-Reply-To: <4758E32C.3010702@ar.media.kyoto-u.ac.jp>
References: <a2b3004b0712060949y6eadd32bha153d89d38d07ac0@mail.gmail.com>	<47583C18.7050100@aero.iitb.ac.in>	<4758C6DE.3060205@ar.media.kyoto-u.ac.jp>	<4758DB8A.9000002@aero.iitb.ac.in>
	<4758E32C.3010702@ar.media.kyoto-u.ac.jp>
Message-ID: <4759073D.4080206@polyu.edu.hk>

David Cournapeau wrote:
> Prabhu Ramachandran wrote:
>> David Cournapeau wrote:
>>   
>>> The whole problem is coming from the fact that Lorenzo installed things 
>>> in /usr. To keep things simple: never ever do that. For example, if you 
>>>     
>> On Debian/debian-derivatives the recommended way is to simply do
>>
>> easy_install --prefix=/usr/local [other args]
>>   
> /usr/local is indeed supposed to be the default path for such things 
> (that's the default path used by GNU build system, for example), and the 
> one where to install things not under the control of your distribution 
> (there is /opt, too; but I don't think it matters much which one you use).
> 
> David

Hello,

A quick and dirty way to make a (non-policy conforming) debian package 
is checkinstall. It's apt-getable.

sid
http://packages.debian.org/sid/checkinstall

Homepage
http://asic-linux.com.mx/~izto/checkinstall/

I've no experience of using it with easy_install though, but it works 
well with the GNU way. Just issue "checkinstall make install" instead of 
"make install" as root and the package will be installed/managed by dpkg.

Regards,
ST
--



From jre at enthought.com  Fri Dec  7 17:57:06 2007
From: jre at enthought.com (J. Ryan Earl)
Date: Fri, 07 Dec 2007 16:57:06 -0600
Subject: [SciPy-user] Scipy.org having problems?
In-Reply-To: <db6b5ecc0712062338g608153f6s789b496709384606@mail.gmail.com>
References: <db6b5ecc0712062321j2aaa40fft1663aa14490d22c8@mail.gmail.com>	<4758F545.9020503@gmail.com>
	<db6b5ecc0712062338g608153f6s789b496709384606@mail.gmail.com>
Message-ID: <4759CFC2.9010603@enthought.com>

Hello,

We are currently in the process of migrating SciPy.org onto better 
hardware.  The current hardware is overloaded, but the new hardware will 
handle around six times as much capacity.

Cheers,
-ryan

Fernando Perez wrote:
> On Dec 7, 2007 12:24 AM, Robert Kern <robert.kern at gmail.com> wrote:
>
>   
>> It's working fine for me, but I've forwarded your message on to Ryan.
>>     
>
> Thanks!
>
> The load average was hovering around 7 a moment ago, with a bunch of
> trac.fcgi processes, in case it helps.
>
> M    TIME+  COMMAND
> 31517 scipy     15   0 50312  19m 4232 S 42.1  1.0   2:08.46 trac.fcgi
> 23756 scipy     15   0 88872  19m 4232 S 23.2  1.0  18:17.42 trac.fcgi
>  1800 ipython   15   0 42316  15m 3952 S 18.9  0.8   0:24.71 trac.fcgi
> 30365 scipy     15   0 42340  14m 4292 S  7.6  0.7   0:29.47 trac.fcgi
> 27390 astropy   16   0 62168  22m 3960 S  6.0  1.1  32:02.84 trac.fcgi
> 24035 scipy     15   0 41784  12m 4144 S  4.6  0.6   0:45.02 trac.fcgi
>
>
> Cheers,
>
> f
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   



From stef.mientki at gmail.com  Sat Dec  8 04:54:20 2007
From: stef.mientki at gmail.com (stef mientki)
Date: Sat, 08 Dec 2007 10:54:20 +0100
Subject: [SciPy-user] Scipy + wxPython ...
Message-ID: <475A69CC.8000307@gmail.com>

hello,

A month ago or so we discussed the possibilities to use Vision as a 
possibility
to enhance Scipy to make it more look like some of the commercial 
scientific packages.

I looked if I could use Vision,
but as some of you already said: "Vision is not very pretty" ,
and I have to admit I didn't get it working within a limited amount of time.
Also the other packages, Orange and Elefant,
were too difficult for me to get them working in a short time.

So therefor I just started to fork another application I'm working on,
which just uses wxPython, to see what I could accomplish.
Although it's just a "proof of concept" ,
made from a quick and dirty glueing of some (most already existing) 
components,
a large part of the program is already working quit well.

If you're interested, look here:
http://stef.mientki.googlepages.com/pylab_works_1.html
(10MB, 8 minutes)

Any comment, suggestions, domain expertise are very welcome.

Hearing that Enthought is working on something similar,
I'm still very curious to see their concept.


cheers,
Stef


From matthieu.brucher at gmail.com  Sat Dec  8 05:18:30 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sat, 8 Dec 2007 11:18:30 +0100
Subject: [SciPy-user] Scipy + wxPython ...
In-Reply-To: <475A69CC.8000307@gmail.com>
References: <475A69CC.8000307@gmail.com>
Message-ID: <e76aa17f0712080218w4a8371dcn6539fed55bad30d3@mail.gmail.com>

2007/12/8, stef mientki <stef.mientki at gmail.com>:
>
> hello,
>
> A month ago or so we discussed the possibilities to use Vision as a
> possibility
> to enhance Scipy to make it more look like some of the commercial
> scientific packages.
>
> I looked if I could use Vision,
> but as some of you already said: "Vision is not very pretty" ,
> and I have to admit I didn't get it working within a limited amount of
> time.
> Also the other packages, Orange and Elefant,
> were too difficult for me to get them working in a short time.
>
> So therefor I just started to fork another application I'm working on,
> which just uses wxPython, to see what I could accomplish.
> Although it's just a "proof of concept" ,
> made from a quick and dirty glueing of some (most already existing)
> components,
> a large part of the program is already working quit well.
>
> If you're interested, look here:
> http://stef.mientki.googlepages.com/pylab_works_1.html
> (10MB, 8 minutes)
>
> Any comment, suggestions, domain expertise are very welcome.



It's great, but the fact that you do real time job as well splits
applications in two : the one that need it and teh one that must not have
it. If I have to iterate on a folder and apply something to each element in
the folder and then save it, I would use a for loop with a sub worflow with
the load, process and save boxes inside the for loop box, not outside.

Now, I think that if an input is connected to something, the associated
editor should not be displayed.

But great job, I didn't create such an application that works that well even
if I started a long time ago ;)


Hearing that Enthought is working on something similar,
> I'm still very curious to see their concept.


Yes, their block_canvas will be a huge competitor with the support of the
Traits library for modifying and/or displaying variables. For what I saw, it
will beat Vision hands down (their workflow is a Python script directly, so
adding new functions is generally straightforward).

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/20071208/355f3417/attachment.html>

From s.mientki at ru.nl  Sat Dec  8 11:02:09 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Sat, 08 Dec 2007 17:02:09 +0100
Subject: [SciPy-user] Scipy + wxPython ...
In-Reply-To: <e76aa17f0712080218w4a8371dcn6539fed55bad30d3@mail.gmail.com>
References: <475A69CC.8000307@gmail.com>
	<e76aa17f0712080218w4a8371dcn6539fed55bad30d3@mail.gmail.com>
Message-ID: <475AC001.1050704@ru.nl>



Matthieu Brucher wrote:
>
>
> 2007/12/8, stef mientki <stef.mientki at gmail.com 
> <mailto:stef.mientki at gmail.com>>:
>
>     hello,
>
>     A month ago or so we discussed the possibilities to use Vision as a
>     possibility
>     to enhance Scipy to make it more look like some of the commercial
>     scientific packages.
>
>     I looked if I could use Vision,
>     but as some of you already said: "Vision is not very pretty" ,
>     and I have to admit I didn't get it working within a limited
>     amount of time.
>     Also the other packages, Orange and Elefant,
>     were too difficult for me to get them working in a short time.
>
>     So therefor I just started to fork another application I'm working on,
>     which just uses wxPython, to see what I could accomplish.
>     Although it's just a "proof of concept" ,
>     made from a quick and dirty glueing of some (most already existing)
>     components,
>     a large part of the program is already working quit well.
>
>     If you're interested, look here:
>     http://stef.mientki.googlepages.com/pylab_works_1.html
>     <http://stef.mientki.googlepages.com/pylab_works_1.html>
>     (10MB, 8 minutes)
>
>     Any comment, suggestions, domain expertise are very welcome.
>
>
>
> It's great, but the fact that you do real time job as well splits 
> applications in two : the one that need it and teh one that must not 
> have it. If I have to iterate on a folder and apply something to each 
> element in the folder and then save it, I would use a for loop with a 
> sub worflow with the load, process and save boxes inside the for loop 
> box, not outside.
Yes, I've no ready solution for that.
Labview uses "time-critical" signals for that.
I was thinking of grouping bricks together, so they form a new brick, in 
which closed for loops could exists.
>
> Now, I think that if an input is connected to something, the 
> associated editor should not be displayed.
Good point, also makes the design more consistent.
I now see also an error in the demo, I throw all the controls of the 
rotation away, while the interpolation should remain.
>
> But great job, I didn't create such an application that works that 
> well even if I started a long time ago ;)
>  
>
>     Hearing that Enthought is working on something similar,
>     I'm still very curious to see their concept. 
>
>
> Yes, their block_canvas will be a huge competitor with the support of 
> the Traits library for modifying and/or displaying variables. For what 
> I saw, it will beat Vision hands down (their workflow is a Python 
> script directly, so adding new functions is generally straightforward).
I'ld love to see some peek previews ;-)
cheers,
Stef



From webb.sprague at gmail.com  Sat Dec  8 19:19:56 2007
From: webb.sprague at gmail.com (Webb Sprague)
Date: Sat, 8 Dec 2007 16:19:56 -0800
Subject: [SciPy-user] Estimating AR(1) model?
Message-ID: <b11ea23c0712081619h638fba46sdc4ebcb6e15d803@mail.gmail.com>

Hi all,

Is there a quick approach for estimating a simple AR(1) time series
model from a vector of data points?  Currently I use lingregress on
lagged data and roll my own standard error computations.  It would be
great if someone smarter than I has done this already, especially if
they use MLE or Yule-Walker.

I am using Gentoo, scipy.__version__ == '0.6.0', and I can't find a
scipy/stats/models directory in my installation, if that matters.

Thx
W


From david at ar.media.kyoto-u.ac.jp  Sat Dec  8 22:23:03 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Sun, 09 Dec 2007 12:23:03 +0900
Subject: [SciPy-user] Estimating AR(1) model?
In-Reply-To: <b11ea23c0712081619h638fba46sdc4ebcb6e15d803@mail.gmail.com>
References: <b11ea23c0712081619h638fba46sdc4ebcb6e15d803@mail.gmail.com>
Message-ID: <475B5F97.1070506@ar.media.kyoto-u.ac.jp>

Webb Sprague wrote:
> Hi all,
>
> Is there a quick approach for estimating a simple AR(1) time series
> model from a vector of data points?  Currently I use lingregress on
> lagged data and roll my own standard error computations.  It would be
> great if someone smarter than I has done this already, especially if
> they use MLE or Yule-Walker.
>
>   
Using Yule-Walker (and something like Levinson Durbin to solve it) for 
AR(1) is kind of overkill. There is a function similar to lpc into my 
sandbox (cdavid: from scipy.sandbox.cdavid import lpc). This gives you 
the ar coefficients, the error, and the k coefficients (follows matlab 
lpc function).

cheers,

David


From odonnems at yahoo.com  Sat Dec  8 23:00:52 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Sat, 8 Dec 2007 20:00:52 -0800 (PST)
Subject: [SciPy-user] Can someone shed light on this simple weave inline test
Message-ID: <716781.78635.qm@web58006.mail.re3.yahoo.com>

I am trying to do some testing and I can not get the following test to work. The code is compiled and the values returned are:
>>> <weave: compiling>
creating c:\temp\Michael\python25_intermediate\compiler_12e837eb1ea3ab5199fbcc0e83015e3f
0 0 0 0 0 0 0 0 0 0

I am able to return an array using a similar method, but this is not working for some reason. 

Is there any documentation that may help me with passing multiple variables and arrays?


Thanks,
Michael


winx = 10
winy = 10
kernel = numpy.random.randint(1,10, (10,10)) #Values between 0 and 10 for an array with 10 x 10 shape
#Now make it floating
#kernel * 1.0
nullValue = -10000
ND=NNW=WNW=WSW=SSW=SSE=ESE=ENE=NNE=Flat=0

code = r"""
        #include "Python.h"
        //int ND, NNW, WNW, WSW, SSW, SSE, ESE, ENE, NNE, Flat;
        
        //Cycle through kernel and calculate the frequency of each compass direction           
        //Change row in kernel
        for (int k=0; k<winy; k++) {
            //Change column in kernel
            for (int l=0; l<winx; l++) {
                if (kernel(k,l) == nullValue)  //or written as numpy.take(kernel, [0]) == [0]
                    ND++;  //Null Data so not evaluating; or written as ND++ in C++
                else if (kernel(k,l) == 1)
                    NNW++;
                else if (kernel(k,l) == 2)
                    WNW++;
                else if (kernel(k,l) == 3)
                    WSW++;
                else if (kernel(k,l) == 4)
                    SSW++;
                else if (kernel(k,l) == 5)
                    SSE++;
                else if (kernel(k,l) == 6)
                    ESE++;
                else if (kernel(k,l) == 7)
                    ENE++;
                else if (kernel(k,l) == 8)
                    NNE++;
                else if (kernel(k,l) == 9)
                    Flat++;
            }
        }
        """
weave.inline(code,['winy', 'winx', 'nullValue', 'kernel', 'ND', 'NNW', 'WNW', 'WSW', 'SSW', 'SSE', 'ESE', 'ENE', 'NNE', 'Flat'], type_converters=converters.blitz, compiler='gcc')

print ND, NNW, WNW, WSW,SSW, SSE, ESE, ENE, NNE, Flat





      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071208/b5045a11/attachment.html>

From steveire at gmail.com  Sun Dec  9 19:41:31 2007
From: steveire at gmail.com (Stephen Kelly)
Date: Mon, 10 Dec 2007 00:41:31 +0000
Subject: [SciPy-user] gmane.comp.kde.devel.core: Authorization required
In-Reply-To: <E1J1WbZ-000511-Mn@ciao.gmane.org>
References: <fji1is$iao$1@ger.gmane.org> <E1J1WbZ-000511-Mn@ciao.gmane.org>
Message-ID: <200712100041.31966.steveire@gmail.com>

On Monday 10 December 2007 00:34:41 Gmane Autoauthorizer wrote:
> You have sent a message to be posted on the
> gmane.comp.kde.devel.core newsgroup.
>
>
> This is a non-public mailing list, which means that you have to
> subscribe to the list to post to it.  If you're already subscribed to
> the list, Gmane can forward the message you sent to the list if you respond
> to this message.  If not, you should sign up to the mailing list first,
> and then respond to this message, or just forget about it.
>
> Many mailing lists have an option to subscribe to a list, but then put
> it in 'nomail' mode, which means that you won't receive any mail from
> the list.
>
> The mailing list software used for the list in question is mailman.
>
>
> First subscribe to the mailing list.  The subscription address is
> kde-core-devel-request at kde.org.  Then send a message to
> kde-core-devel-request at kde.org with the text
>
>   set nomail on YOUR_PASSWORD_HERE
>
> or (depending on the version of Mailman)
>
>   set authenticate YOUR_PASSWORD_HERE
>   set delivery off
>
> in the body of the message.
>
> You have to respond within one week.




From cimrman3 at ntc.zcu.cz  Mon Dec 10 04:53:27 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Mon, 10 Dec 2007 10:53:27 +0100
Subject: [SciPy-user] Sparse eigensolver
In-Reply-To: <80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
References: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
	<80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
Message-ID: <475D0C97.2010108@ntc.zcu.cz>


There is also scipy.sandbox.lobpcg (pure scipy code, depending 
temporarily on symeig, until symeig functionality is in scipy)

It seems to run with recent scipy, but lobpcg/tests/benchmark.py gives 
two additional zero eigenvalues that are not in symeig results. Nils, 
have you tried it recently? This used to work...

r.

lorenzo bolla wrote:
> sure: scipy.sandbox.arpack
> it's in the sandbox, so you have to enable it and recompile scipy.
> note that there is a patch for arpack.py (google for it).
> L.
> 
> 
> On 12/6/07, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:
>> Hi,
>>
>> Is there a package for scipy (or a function in scipy directly) that can
>> compute the eigen values and the eigen vectors of a sparse matrix, like the
>> eigs function in Malab ?


From lbolla at gmail.com  Mon Dec 10 08:37:16 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Mon, 10 Dec 2007 14:37:16 +0100
Subject: [SciPy-user] Sparse eigensolver
In-Reply-To: <e76aa17f0712061123s69d1231bqce047126dbbe94c7@mail.gmail.com>
References: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
	<80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
	<e76aa17f0712060901w36ad2e63s7312f59c8a36738e@mail.gmail.com>
	<80c99e790712060955i21fa05f5w1e30957ca4ec6193@mail.gmail.com>
	<e76aa17f0712061123s69d1231bqce047126dbbe94c7@mail.gmail.com>
Message-ID: <80c99e790712100537r725f5046k293b5ee22ab76567@mail.gmail.com>

that would be great!
L.


On 12/6/07, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:
>
> Well, if I need it (I can use a dense one for the moment) and if people
> agree, I'll make it a scikit ;)
>
> Matthieu
>
> 2007/12/6, lorenzo bolla < lbolla at gmail.com>:
> >
> > I agree...
> >
> > On Dec 6, 2007 6:01 PM, Matthieu Brucher <matthieu.brucher at gmail.com>
> > wrote:
> >
> > > Thank you for this. I hope it will be put in the trunk at some point
> > > (or in a scikit) as it would be needed for a manifold learning scikit.
> > >
> > > Matthieu
> > >
> > > 2007/12/6, lorenzo bolla < lbolla at gmail.com>:
> > > >
> > > > sure: scipy.sandbox.arpack
> > > > it's in the sandbox, so you have to enable it and recompile scipy.
> > > > note that there is a patch for arpack.py (google for it).
> > > > L.
> > > >
> > > >
> > > >  On 12/6/07, Matthieu Brucher < matthieu.brucher at gmail.com > wrote:
> > > >
> > > > > Hi,
> > > > >
> > > > > Is there a package for scipy (or a function in scipy directly)
> > > > > that can compute the eigen values and the eigen vectors of a sparse matrix,
> > > > > like the eigs function in Malab ?
> > > > >
> > > > > 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
> > > > >
> > > > >
> > > >
> > > > _______________________________________________
> > > > 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
> > >
> > > _______________________________________________
> > > 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
> >
> >
>
>
> --
> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071210/93f53dd9/attachment.html>

From nwagner at iam.uni-stuttgart.de  Mon Dec 10 12:03:11 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Mon, 10 Dec 2007 18:03:11 +0100
Subject: [SciPy-user] Sparse eigensolver
In-Reply-To: <475D0C97.2010108@ntc.zcu.cz>
References: <e76aa17f0712060735w3a0d3f19n27c3842f99810873@mail.gmail.com>
	<80c99e790712060852r3a879ef5vf814ab78514e94ce@mail.gmail.com>
	<475D0C97.2010108@ntc.zcu.cz>
Message-ID: <web-93147137@uni-stuttgart.de>

On Mon, 10 Dec 2007 10:53:27 +0100
  Robert Cimrman <cimrman3 at ntc.zcu.cz> wrote:
> 
> There is also scipy.sandbox.lobpcg (pure scipy code, 
>depending 
> temporarily on symeig, until symeig functionality is in 
>scipy)
> 
> It seems to run with recent scipy, but 
>lobpcg/tests/benchmark.py gives 
> two additional zero eigenvalues that are not in symeig 
>results. Nils, 
> have you tried it recently? This used to work...
  
python -i 
local/lib64/python2.3/site-packages/scipy/sandbox/lobpcg/tests/benchmark.py
****** 128

Results by LOBPCG

128 [   1.    4.    9.   16.   25.   36.   49.   64.   81. 
 100.]

Results by symeig

128 [  1.   4.   9.  16.  25.  36.  49.  64.  81. 100.]
****** 256

Results by LOBPCG

256 [   1.    4.    9.   16.   25.   36.   49.   64.   81. 
 100.]

Results by symeig

256 [  1.   4.   9.  16.  25.  36.  49.  64.  81. 100.]
****** 512

Results by LOBPCG

512 [  1.   4.   9.  16.  25.  36.  49.  64.  81. 100.]

Results by symeig

512 [   1.    4.    9.   16.   25.   36.   49.   64.   81. 
 100.]
****** 1024

Results by LOBPCG

1024 [   1.    4.    9.   16.   25.   36.   49.   64. 
  81.  100.]

Results by symeig

1024 [  1.   4.   9.  16.  25.  36.  49.  64.  81. 100.]
****** 2048

Results by LOBPCG

2048 [  1.   4.   9.  16.  25.  36.  49.  64.  81. 100.]

Results by symeig

2048 [   1.    4.    9.   16.   25.   36.   49.   64. 
  81.  100.]
  
Cheers,

              Nils


From nwagner at iam.uni-stuttgart.de  Mon Dec 10 13:43:32 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Mon, 10 Dec 2007 19:43:32 +0100
Subject: [SciPy-user] Slicing arrays
Message-ID: <web-93150492@uni-stuttgart.de>

Hi all,

I have an array

>>> a
array([ 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, 50,
        51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 
63, 64, 65])

How can I extract an array b from a such that

>>> b
array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24, 
25, 26, 30, 31,
        32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55, 
56, 60, 61, 62])

Nils


From odonnems at yahoo.com  Mon Dec 10 13:47:36 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Mon, 10 Dec 2007 10:47:36 -0800 (PST)
Subject: [SciPy-user] using STL in scipy weave inline
Message-ID: <124989.33872.qm@web58006.mail.re3.yahoo.com>

OS Microsoft XP SP2 x86
python 2.5.1
compiler: mingw

Does anyone have experience using multimap or the standard template library within weave inline? I need to create a map that contains numeric keys which are paired to a string, but I get an error when declaring the multimap function. If i include <map> I get other errors about namespace and if don't then I get an error for not declaring the multimap. Neither method works.

TIA,
Michael

Example of code:

total = (winx * winy * 1.0) #Total number of cells to evaluate in kernel

code = r"""
            double ND, NNW, WNW, WSW,SSW, SSE, ESE, ENE, NNE, Flat;
            ND = 22.0;
            NNW = 11.0;
            WNW = 10.0;
            WSW = 30.0;
            SSW = 8.0;
            SSE = 2.0;
            ESE = 1.0;
            ENE = 5.0;
            NNE = 5.0;
            Flat = 3.0;

            //Calculate the frequency of the counts
            fND = (ND / total);    
            fNNW = (NNW / total);   
            fWNW = (WNW / total);  
            fWSW = (WSW / total);  
            fSSW = (SSW / total);   
            fSSE = (SSE / total);   
            fESE = (ESE / total);  
            fENE = (ENE / total);  
            fNNE = (NNE / total);   
            fFlat = (Flat / total);  
        
            //Create a map for pairs: frequency, string
            multimap<double, string> map_freq; //Create empty map
        
            map_freq[fND) = 'fND';
            map_freq[fNNW] = 'fNNW';
            map_freq[fWNW] = 'fWNW';
            map_freq[fWSW] = 'fWSW';
            map_freq[fSSW] = 'fSSW';
            map_freq[fSSE] = 'fSSE';
            map_freq[fESE] = 'fESE';
            map_freq[fENE] = 'fENE';
            map_freq[fNNE] = 'fNNE';
            map_freq[fFlat] = 'fFlat';
             """

weave.inline(code,['total'], type_converters=converters.blitz, compiler='gcc')




      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071210/b8e983c7/attachment.html>

From robert.kern at gmail.com  Mon Dec 10 14:58:25 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 10 Dec 2007 13:58:25 -0600
Subject: [SciPy-user] Slicing arrays
In-Reply-To: <web-93150492@uni-stuttgart.de>
References: <web-93150492@uni-stuttgart.de>
Message-ID: <475D9A61.2040607@gmail.com>

Nils Wagner wrote:
> Hi all,
> 
> I have an array
> 
>>>> a
> array([ 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, 50,
>         51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 
> 63, 64, 65])
> 
> How can I extract an array b from a such that
> 
>>>> b
> array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24, 
> 25, 26, 30, 31,
>         32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55, 
> 56, 60, 61, 62])

In [1]: from numpy import *

In [2]: a = arange(66)

In [3]: b = reshape(a, (-1, 6))[:,:3].ravel()

In [4]: b
Out[4]:
array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24, 25, 26, 30, 31,
       32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55, 56, 60, 61, 62])

-- 
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 gael.varoquaux at normalesup.org  Mon Dec 10 15:14:41 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 10 Dec 2007 21:14:41 +0100
Subject: [SciPy-user] Slicing arrays
In-Reply-To: <web-93150492@uni-stuttgart.de>
References: <web-93150492@uni-stuttgart.de>
Message-ID: <20071210201441.GA25289@clipper.ens.fr>

On Mon, Dec 10, 2007 at 07:43:32PM +0100, Nils Wagner wrote:
> I have an array

> >>> a
> array([ 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, 50,
>         51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 
> 63, 64, 65])

> How can I extract an array b from a such that

> >>> b
> array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24, 
> 25, 26, 30, 31,
>         32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55, 
> 56, 60, 61, 62])

I don't understand what is the relationship between a and b.

Ga?l


From Karl.Young at ucsf.edu  Mon Dec 10 14:59:22 2007
From: Karl.Young at ucsf.edu (Karl Young)
Date: Mon, 10 Dec 2007 11:59:22 -0800
Subject: [SciPy-user] Slicing arrays
In-Reply-To: <web-93150492@uni-stuttgart.de>
References: <web-93150492@uni-stuttgart.de>
Message-ID: <475D9A9A.2060708@ucsf.edu>


If you don't care about elegance or efficiency how about:

--------------------------------------------------------------------------------------------------
import scipy

a = scipy.array([ 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, 
50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65])

runs = 8
space = 4

b = scipy.array(a[1:runs+1])
for i in range(1,(len(a)/runs)-1):
    b = scipy.concatenate((b,a[b[-1]+space:b[-1]+runs+space+1]),axis=1)
-------------------------------------------------------------------------------------------------

PS If you really meant to cut off the last three elements of a, i'll 
leave that step to you...

>Hi all,
>
>I have an array
>
>  
>
>>>>a
>>>>        
>>>>
>array([ 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, 50,
>        51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 
>63, 64, 65])
>
>How can I extract an array b from a such that
>
>  
>
>>>>b
>>>>        
>>>>
>array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24, 
>25, 26, 30, 31,
>        32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55, 
>56, 60, 61, 62])
>
>Nils
>_______________________________________________
>SciPy-user mailing list
>SciPy-user at scipy.org
>http://projects.scipy.org/mailman/listinfo/scipy-user
>
>  
>


-- 

Karl Young
Center for Imaging of Neurodegenerative Diseases, UCSF          
VA Medical Center (114M)              Phone:  (415) 221-4810 x3114  lab        
4150 Clement Street                   FAX:    (415) 668-2864
San Francisco, CA 94121               Email:  karl young at ucsf edu


From vincefn at users.sourceforge.net  Mon Dec 10 15:46:01 2007
From: vincefn at users.sourceforge.net (Vincent Favre-Nicolin)
Date: Mon, 10 Dec 2007 21:46:01 +0100
Subject: [SciPy-user] using STL in scipy weave inline
In-Reply-To: <529E0C005F46104BA9DB3CB93F3979750117D4C8@TOKYO.intra.cea.fr>
References: <529E0C005F46104BA9DB3CB93F3979750117D4C8@TOKYO.intra.cea.fr>
Message-ID: <200712102146.01404.vincefn@users.sourceforge.net>

On lundi 10 d?cembre 2007, Michael ODonnell wrote:
> OS Microsoft XP SP2 x86
> python 2.5.1
> compiler: mingw
>
> Does anyone have experience using multimap or the standard template library
> within weave inline? I need to create a map that contains numeric keys
> which are paired to a string, but I get an error when declaring the
> multimap function. If i include <map> I get other errors about namespace
> and if don't then I get an error for not declaring the multimap. Neither
> method works.

   I have not used the STL with weave, but have you tried either :

(1) adding "using namespace std;" to the beginning of your code

(2) using "std::multimap" (and same for string : "std::string") for all the 
STL classes you use ?

-- 
Vincent Favre-Nicolin
Universit? Joseph Fourier
http://v.favrenicolin.free.fr
ObjCryst & Fox : http://objcryst.sourceforge.net


From tom.denniston at alum.dartmouth.org  Mon Dec 10 15:49:12 2007
From: tom.denniston at alum.dartmouth.org (Tom Denniston)
Date: Mon, 10 Dec 2007 14:49:12 -0600
Subject: [SciPy-user] Slicing arrays
In-Reply-To: <web-93150492@uni-stuttgart.de>
References: <web-93150492@uni-stuttgart.de>
Message-ID: <d6f2d3dd0712101249x19831b47jc5e42d4ab2fffb75@mail.gmail.com>

Something like this should work:

In [8]: a1 = numpy.ones(100)

In [9]: a1[::3]=3

In [11]: numpy.cumsum(a1)-2

array([   1.,    2.,    3.,    6.,    7.,    8.,   11.,   12.,   13.,
         16.,   17.,   18.,   21.,   22.,   23.,   26.,   27.,   28.,
         31.,   32.,   33.,   36.,   37.,   38.,   41.,   42.,   43.,
         46.,   47.,   48.,   51.,   52.,   53.,   56.,   57.,   58.,
         61.,   62.,   63.,   66.,   67.,   68.,   71.,   72.,   73.,
         76.,   77.,   78.,   81.,   82.,   83.,   86.,   87.,   88.,
         91.,   92.,   93.,   96.,   97.,   98.,  101.,  102.,  103.,
        106.,  107.,  108.,  111.,  112.,  113.,  116.,  117.,  118.,
        121.,  122.,  123.,  126.,  127.,  128.,  131.,  132.,  133.,
        136.,  137.,  138.,  141.,  142.,  143.,  146.,  147.,  148.,
        151.,  152.,  153.,  156.,  157.,  158.,  161.,  162.,  163.,  166.])


On 12/10/07, Nils Wagner <nwagner at iam.uni-stuttgart.de> wrote:
> Hi all,
>
> I have an array
>
> >>> a
> array([ 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, 50,
>        51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62,
> 63, 64, 65])
>
> How can I extract an array b from a such that
>
> >>> b
> array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24,
> 25, 26, 30, 31,
>        32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55,
> 56, 60, 61, 62])
>
> Nils
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From as8ca at virginia.edu  Mon Dec 10 15:57:07 2007
From: as8ca at virginia.edu (Alok Singhal)
Date: Mon, 10 Dec 2007 15:57:07 -0500
Subject: [SciPy-user] Slicing arrays
In-Reply-To: <web-93150492@uni-stuttgart.de>
References: <web-93150492@uni-stuttgart.de>
Message-ID: <20071210205707.GH767@virginia.edu>

On 10/12/07: 19:43, Nils Wagner wrote:
> Hi all,
> 
> I have an array
> 
> >>> a
> array([ 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, 50,
>         51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 
> 63, 64, 65])
> 
> How can I extract an array b from a such that
> 
> >>> b
> array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24, 
> 25, 26, 30, 31,
>         32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55, 
> 56, 60, 61, 62])

I can do it by:

b = a.copy()
b.shape = (-1,6)
b = b[:,0:3].copy()
b.shape = (1, -1)
b = b.squeeze()

Maybe there is a more efficient method, but this seems to work.

-Alok

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


From silva at lma.cnrs-mrs.fr  Mon Dec 10 17:07:59 2007
From: silva at lma.cnrs-mrs.fr (Fabrice Silva)
Date: Mon, 10 Dec 2007 23:07:59 +0100
Subject: [SciPy-user] Slicing arrays
In-Reply-To: <web-93150492@uni-stuttgart.de>
References: <web-93150492@uni-stuttgart.de>
Message-ID: <1197324480.3220.1.camel@localhost.localdomain>

>>> a=arange(66)
>>> a
array([ 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, 50,
       51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65])
>>> tmp = a.reshape(6,11)
>>> tmp
array([[ 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, 50, 51, 52, 53, 54],
       [55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65]])
>>> b=tmp[:,0:3]
>>> b
array([[ 0,  1,  2],
       [ 6,  7,  8],
       [12, 13, 14],
       [18, 19, 20],
       [24, 25, 26],
       [30, 31, 32],
       [36, 37, 38],
       [42, 43, 44],
       [48, 49, 50],
       [54, 55, 56],
       [60, 61, 62]])
>>> b.ravel()
array([ 0,  1,  2,  6,  7,  8, 12, 13, 14, 18, 19, 20, 24, 25, 26, 30, 31,
       32, 36, 37, 38, 42, 43, 44, 48, 49, 50, 54, 55, 56, 60, 61, 62])

-- 
Fabrice Silva <silva at lma.cnrs-mrs.fr>
LMA UPR CNRS 7051 - ?quipe S2M



From eric at enthought.com  Mon Dec 10 18:17:06 2007
From: eric at enthought.com (eric jones)
Date: Mon, 10 Dec 2007 17:17:06 -0600
Subject: [SciPy-user] using STL in scipy weave inline
In-Reply-To: <124989.33872.qm@web58006.mail.re3.yahoo.com>
References: <124989.33872.qm@web58006.mail.re3.yahoo.com>
Message-ID: <475DC8F2.8040705@enthought.com>

Michael ODonnell wrote:
> OS Microsoft XP SP2 x86
> python 2.5.1
> compiler: mingw
>
> Does anyone have experience using multimap or the standard template 
> library within weave inline? I need to create a map that contains 
> numeric keys which are paired to a string, but I get an error when 
> declaring the multimap function. If i include <map> I get other errors 
> about namespace and if don't then I get an error for not declaring the 
> multimap. Neither method works.
There is no problem using the stl from weave.  In fact, it uses 
std::string internally.

You can use

    headers=["<map>"]

as a keyword to the inline call to get map into the namespace.  Once you 
do that, you'll need to either declare "using namespace std;" at the top 
of your code or use std::multimap and std::string to refer to the 
classes in the stl.  Once you've done that, you'll still have some 
issues as multimap doesn't support the operator [], and your C++ strings 
have single quotes around them instead of double quotes.

I've attached an example for your code that instead works using a 
standard python dictionary (use py::dict within weave).  The output is 
shown here:

In [33]: run we_ex.py
<weave: compiling>

In [34]: res
Out[34]:
{0.050000000000000003: 'fESE',
 0.10000000000000001: 'fSSE',
 0.14999999999999999: 'fFlat',
 0.25: 'fNNE',
 0.40000000000000002: 'fSSW',
 0.5: 'fWNW',
 0.55000000000000004: 'fNNW',
 1.1000000000000001: 'fND',
 1.5: 'fWSW'}

If you need to use the stl for some other reason, then refer to this to 
see how to do it: http://www.sgi.com/tech/stl/Multimap.html

Good luck,
eric
>
> TIA,
> Michael
>
> Example of code:
>
> total = (winx * winy * 1.0) #Total number of cells to evaluate in kernel
>
> code = r"""
>             double ND, NNW, WNW, WSW,SSW, SSE, ESE, ENE, NNE, Flat;
>             ND = 22.0;
>             NNW = 11.0;
>             WNW = 10.0;
>             WSW = 30.0;
>             SSW = 8.0;
>             SSE = 2.0;
>             ESE = 1.0;
>             ENE = 5.0;
>             NNE = 5.0;
>             Flat = 3.0;
>
>             //Calculate the frequency of the counts
>             fND = (ND / total);   
>             fNNW = (NNW / total);  
>             fWNW = (WNW / total); 
>             fWSW = (WSW / total); 
>             fSSW = (SSW / total);  
>             fSSE = (SSE / total);  
>             fESE = (ESE / total); 
>             fENE = (ENE / total); 
>             fNNE = (NNE / total);  
>             fFlat = (Flat / total); 
>        
>             //Create a map for pairs: frequency, string
>             multimap<double, string> map_freq; //Create empty map
>        
>             map_freq[fND) = 'fND';
>             map_freq[fNNW] = 'fNNW';
>             map_freq[fWNW] = 'fWNW';
>             map_freq[fWSW] = 'fWSW';
>             map_freq[fSSW] = 'fSSW';
>             map_freq[fSSE] = 'fSSE';
>             map_freq[fESE] = 'fESE';
>             map_freq[fENE] = 'fENE';
>             map_freq[fNNE] = 'fNNE';
>             map_freq[fFlat] = 'fFlat';
>             """
>
> weave.inline(code,['total'], type_converters=converters.blitz, 
> compiler='gcc')
>
> ------------------------------------------------------------------------
> Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try 
> it now. 
> <http://us.rd.yahoo.com/evt=51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ%20> 
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: we_ex.py
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071210/992ed42f/attachment.ksh>

From nmelgarejodiaz at gmail.com  Tue Dec 11 06:04:16 2007
From: nmelgarejodiaz at gmail.com (Natali Melgarejo Diaz)
Date: Tue, 11 Dec 2007 12:04:16 +0100
Subject: [SciPy-user] Green's function
Message-ID: <c2284ea70712110304v2f7d2dcfo89504c97666b6d7d@mail.gmail.com>

Hi to everyone i'm new in this list...i have a question about a
formula i haven't found yet..., do you know the "Green function"??. Is
there anything similar in Scipy??, 'cause in Matllab there is one but
i haven't found it in Scipy yet.

Thanks in advance to everyone ;))


********Natali********


From nikolaskaralis at gmail.com  Tue Dec 11 07:34:07 2007
From: nikolaskaralis at gmail.com (Nikolas Karalis)
Date: Tue, 11 Dec 2007 14:34:07 +0200
Subject: [SciPy-user] Characteristic polynomial
Message-ID: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>

Hello... I'm trying to compute the characteristic polynomial of an integer
(numpy) matrix. But i cannot find any way of doing this.
Do you have any ideas/suggestions?

-- 
Nikolas Karalis
Applied Mathematics and Physics Undergraduate
National Technical University of Athens, Greece
http://users.ntua.gr/ge04042
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071211/47b54222/attachment.html>

From bryanv at enthought.com  Tue Dec 11 10:37:37 2007
From: bryanv at enthought.com (Bryan Van de Ven)
Date: Tue, 11 Dec 2007 09:37:37 -0600
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>
Message-ID: <475EAEC1.4020501@enthought.com>

In [16]: numpy.lib.poly?
Type:           function
Base Class:     <type 'function'>
String Form:    <function poly at 0xb7a66224>
Namespace:      Interactive
File: 
/workspace/python2.5/lib/python2.5/site-packages/numpy/lib/polynomial.py
Definition:     numpy.lib.poly(seq_of_zeros)
Docstring:
     Return a sequence representing a polynomial given a sequence of roots.

     If the input is a matrix, return the characteristic polynomial.

     Example:

         >>> b = roots([1,3,1,5,6])
         >>> poly(b)
         array([ 1.,  3.,  1.,  5.,  6.])


In [17]: a = array([[1,0],[0,2]])

In [18]: coeffs = numpy.lib.poly(a)

In [19]: numpy.lib.poly1d(coeffs)
Out[19]: poly1d([ 1., -3.,  2.])




Nikolas Karalis wrote:
> Hello... I'm trying to compute the characteristic polynomial of an 
> integer (numpy) matrix. But i cannot find any way of doing this.
> Do you have any ideas/suggestions?
> 
> -- 
> Nikolas Karalis
> Applied Mathematics and Physics Undergraduate
> National Technical University of Athens, Greece
> http://users.ntua.gr/ge04042
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From nikolaskaralis at gmail.com  Tue Dec 11 11:38:17 2007
From: nikolaskaralis at gmail.com (Nikolas Karalis)
Date: Tue, 11 Dec 2007 18:38:17 +0200
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <475EAEC1.4020501@enthought.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>
	<475EAEC1.4020501@enthought.com>
Message-ID: <85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>

Hello Bryan and thanks for responding.
I have already tried that, but check this example...

>>> from numpy import *
>>> a=arange(25)
>>> a.shape=5,5
>>> a
array([[ 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]])
>>> coeffs=poly(a)
>>> coeffs
array([  1.00000000e+00,  -6.00000000e+01,  -2.50000000e+02,
        -1.67542241e-13,   1.82969746e-28,  -8.57128973e-45])
>>> poly1d(coeffs)
poly1d([  1.00000000e+00,  -6.00000000e+01,  -2.50000000e+02,
        -1.67542241e-13,   1.82969746e-28,  -8.57128973e-45])

Check these coefficients, like -1.67542241e-13 etc... these numbers are
supposed to be zero, but they are "very small".

The very same example in sage :
A = matrix(5,5, range(25)); A
A.charpoly()

x^5 - 60*x^4 - 250*x^3

I tried to apply round function on the coefficients :
>>> for i in range(len(coeffs)): coeffs[i]=round(coeffs[i])
...
>>> coeffs
array([   1.,  -60., -250.,    0.,    0.,    0.])

and while in this case it returns the right result, there are cases where
the coefficients are like : 1.1234566789e64,
where the results cannot be rounded (since they are already integers) but
they are different where they were supposed to be the same.
A strange example :

For the 10x10 matrices :
A=matrix([[36,8,8,1,1,1,1,1,8,1],
             [8,36,8,8,1,1,1,1,1,1],
             [8,8,25,8,1,8,1,1,1,1],
             [1,8,8,36,8,1,1,1,1,1],
             [1,1,1,8,36,1,8,8,1,1],
             [1,1,8,1,1,36,8,1,1,8],
             [1,1,1,1,8,8,49,1,1,1],
             [1,1,1,1,8,1,1,49,8,1],
             [8,1,1,1,1,1,1,8,36,8],
             [1,1,1,1,1,8,1,1,8,49]])

B=matrix([[36,8,8,1,1,1,1,8,1,1],
             [8,25,8,1,8,8,1,1,1,1],
             [8,8,36,8,1,1,1,1,1,1],
             [1,1,8,36,8,1,1,1,1,8],
             [1,8,1,8,36,8,1,1,1,1],
             [1,8,1,1,8,36,8,1,1,1],
             [1,1,1,1,1,8,49,1,8,1],
             [8,1,1,1,1,1,1,49,8,1],
             [1,1,1,1,1,1,8,8,36,8],
             [1,1,1,8,1,1,1,1,8,49]])

>>> poly(A)
array([  1.00000000e+00,  -3.88000000e+02,   6.65430000e+04,
        -6.63946000e+06,   4.26558031e+08,  -1.84256082e+10,
         5.41543001e+11,  -1.06843860e+13,   1.35292436e+14,
        -9.91717494e+14,   3.19104663e+15])
>>> poly(B)
array([  1.00000000e+00,  -3.88000000e+02,   6.65430000e+04,
        -6.63946000e+06,   4.26558031e+08,  -1.84256082e+10,
         5.41543001e+11,  -1.06843860e+13,   1.35292436e+14,
        -9.91717494e+14,   3.19104663e+15])

sage returns :

x^10 - 388*x^9 + 66543*x^8 - 6639460*x^7 + 426558031*x^6 -
18425608218*x^5 + 541543001174*x^4 - 10684385967388*x^3 +
135292435663763*x^2 - 991717494497240*x + 3191046625350150

for both matrices.

You can see the difference between 3.19104663e+15  and 3191046625350150
On the other hand, if we get the exact value of 3.19104663e+15 (by using :
poly(A)[-1] and poly(B)[-1]) we have :
3191046625350152.0
3191046625350153.5

And with all these... :S I'm really confused.
Anyway, it would be really useful if there existed a function to compute the
modular characteristic polynomial (modulo a big prime) (like in maple).

On Dec 11, 2007 5:37 PM, Bryan Van de Ven <bryanv at enthought.com> wrote:

> In [16]: numpy.lib.poly?
> Type:           function
> Base Class:     <type 'function'>
> String Form:    <function poly at 0xb7a66224>
> Namespace:      Interactive
> File:
> /workspace/python2.5/lib/python2.5/site-packages/numpy/lib/polynomial.py
> Definition:     numpy.lib.poly(seq_of_zeros)
> Docstring:
>     Return a sequence representing a polynomial given a sequence of roots.
>
>     If the input is a matrix, return the characteristic polynomial.
>
>     Example:
>
>         >>> b = roots([1,3,1,5,6])
>         >>> poly(b)
>         array([ 1.,  3.,  1.,  5.,  6.])
>
>
> In [17]: a = array([[1,0],[0,2]])
>
> In [18]: coeffs = numpy.lib.poly(a)
>
> In [19]: numpy.lib.poly1d(coeffs)
> Out[19]: poly1d([ 1., -3.,  2.])
>
>
>
>
> Nikolas Karalis wrote:
> > Hello... I'm trying to compute the characteristic polynomial of an
> > integer (numpy) matrix. But i cannot find any way of doing this.
> > Do you have any ideas/suggestions?
> >
> > --
> > Nikolas Karalis
> > Applied Mathematics and Physics Undergraduate
> > National Technical University of Athens, Greece
> > http://users.ntua.gr/ge04042
> >
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > 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
>



-- 
Nikolas Karalis
Applied Mathematics and Physics Undergraduate
National Technical University of Athens, Greece
http://users.ntua.gr/ge04042
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071211/6055dbfb/attachment.html>

From matthieu.brucher at gmail.com  Tue Dec 11 11:45:03 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Tue, 11 Dec 2007 17:45:03 +0100
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>
	<475EAEC1.4020501@enthought.com>
	<85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>
Message-ID: <e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>

You can't expect scipy to find the exact coefficients. sage is a symbolic
package, it will find the correct answers, but scipy will find only an
approximated one, up to the machine precision. This is what you see in your
exemple.
If you have integers, you could expect scipy to return long integers (exact
result), but this is not the case as almost everything is converted into a
float array before the actual C or Fortran routine is run.

Matthieu

2007/12/11, Nikolas Karalis <nikolaskaralis at gmail.com>:
>
> Hello Bryan and thanks for responding.
> I have already tried that, but check this example...
>
> >>> from numpy import *
> >>> a=arange(25)
> >>> a.shape=5,5
> >>> a
> array([[ 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]])
> >>> coeffs=poly(a)
> >>> coeffs
> array([  1.00000000e+00,  -6.00000000e+01 ,  -2.50000000e+02,
>         -1.67542241e-13,   1.82969746e-28,  -8.57128973e-45])
> >>> poly1d(coeffs)
> poly1d([  1.00000000e+00,  -6.00000000e+01,  -2.50000000e+02,
>         -1.67542241e-13,   1.82969746e-28 ,  -8.57128973e-45])
>
> Check these coefficients, like -1.67542241e-13 etc... these numbers are
> supposed to be zero, but they are "very small".
>
> The very same example in sage :
> A = matrix(5,5, range(25)); A
> A.charpoly()
>
> x^5 - 60*x^4 - 250*x^3
>
> I tried to apply round function on the coefficients :
> >>> for i in range(len(coeffs)): coeffs[i]=round(coeffs[i])
> ...
> >>> coeffs
> array([   1.,  -60., -250.,    0.,    0.,    0.])
>
> and while in this case it returns the right result, there are cases where
> the coefficients are like : 1.1234566789e64,
> where the results cannot be rounded (since they are already integers) but
> they are different where they were supposed to be the same.
> A strange example :
>
> For the 10x10 matrices :
> A=matrix([[36,8,8,1,1,1,1,1,8,1],
>              [8,36,8,8,1,1,1,1,1,1],
>              [8,8,25,8,1,8,1,1,1,1],
>              [1,8,8,36,8,1,1,1,1,1],
>              [1,1,1,8,36,1,8,8,1,1],
>              [1,1,8,1,1,36,8,1,1,8],
>              [1,1,1,1,8,8,49,1,1,1],
>              [1,1,1,1,8,1,1,49,8,1],
>              [8,1,1,1,1,1,1,8,36,8],
>              [1,1,1,1,1,8,1,1,8,49]])
>
> B=matrix([[36,8,8,1,1,1,1,8,1,1],
>              [8,25,8,1,8,8,1,1,1,1],
>              [8,8,36,8,1,1,1,1,1,1],
>              [1,1,8,36,8,1,1,1,1,8],
>              [1,8,1,8,36,8,1,1,1,1],
>              [1,8,1,1,8,36,8,1,1,1],
>              [1,1,1,1,1,8,49,1,8,1],
>              [8,1,1,1,1,1,1,49,8,1],
>              [1,1,1,1,1,1,8,8,36,8],
>              [1,1,1,8,1,1,1,1,8,49]])
>
> >>> poly(A)
> array([  1.00000000e+00,  -3.88000000e+02,   6.65430000e+04,
>         -6.63946000e+06,   4.26558031e+08,  -1.84256082e+10,
>          5.41543001e+11 ,  -1.06843860e+13,   1.35292436e+14,
>         -9.91717494e+14,   3.19104663e+15])
> >>> poly(B)
> array([  1.00000000e+00,  -3.88000000e+02,   6.65430000e+04,
>         -6.63946000e+06,   4.26558031e+08,  - 1.84256082e+10,
>          5.41543001e+11,  -1.06843860e+13,   1.35292436e+14,
>         -9.91717494e+14,   3.19104663e+15])
>
> sage returns :
>
> x^10 - 388*x^9 + 66543*x^8 - 6639460*x^7 + 426558031*x^6 -
>
> 18425608218*x^5 + 541543001174*x^4 - 10684385967388*x^3 +
> 135292435663763*x^2 - 991717494497240*x + 3191046625350150
>
> for both matrices.
>
> You can see the difference between 3.19104663e+15  and 3191046625350150
> On the other hand, if we get the exact value of 3.19104663e+15 (by using :
> poly(A)[-1] and poly(B)[-1]) we have :
> 3191046625350152.0
> 3191046625350153.5
>
> And with all these... :S I'm really confused.
> Anyway, it would be really useful if there existed a function to compute
> the modular characteristic polynomial (modulo a big prime) (like in maple).
>
> On Dec 11, 2007 5:37 PM, Bryan Van de Ven < bryanv at enthought.com> wrote:
>
> > In [16]: numpy.lib.poly ?
> > Type:           function
> > Base Class:     <type 'function'>
> > String Form:    <function poly at 0xb7a66224>
> > Namespace:      Interactive
> > File:
> > /workspace/python2.5/lib/python2.5/site-packages/numpy/lib/polynomial.py
> >
> > Definition:     numpy.lib.poly(seq_of_zeros)
> > Docstring:
> >     Return a sequence representing a polynomial given a sequence of
> > roots.
> >
> >     If the input is a matrix, return the characteristic polynomial.
> >
> >     Example:
> >
> >         >>> b = roots([1,3,1,5,6])
> >         >>> poly(b)
> >         array([ 1.,  3.,  1.,  5.,  6.])
> >
> >
> > In [17]: a = array([[1,0],[0,2]])
> >
> > In [18]: coeffs = numpy.lib.poly (a)
> >
> > In [19]: numpy.lib.poly1d(coeffs)
> > Out[19]: poly1d([ 1., -3.,  2.])
> >
> >
> >
> >
> > Nikolas Karalis wrote:
> > > Hello... I'm trying to compute the characteristic polynomial of an
> > > integer (numpy) matrix. But i cannot find any way of doing this.
> > > Do you have any ideas/suggestions?
> > >
> > > --
> > > Nikolas Karalis
> > > Applied Mathematics and Physics Undergraduate
> > > National Technical University of Athens, Greece
> > > http://users.ntua.gr/ge04042
> > >
> > >
> > >
> > ------------------------------------------------------------------------
> > >
> > > _______________________________________________
> > > 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
> >
>
>
>
> --
> Nikolas Karalis
> Applied Mathematics and Physics Undergraduate
> National Technical University of Athens, Greece
> http://users.ntua.gr/ge04042
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


-- 
French PhD student
Website : http://matthieu-brucher.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/20071211/4bb02224/attachment.html>

From lou_boog2000 at yahoo.com  Tue Dec 11 11:52:19 2007
From: lou_boog2000 at yahoo.com (Lou Pecora)
Date: Tue, 11 Dec 2007 08:52:19 -0800 (PST)
Subject: [SciPy-user] Green's function
In-Reply-To: <c2284ea70712110304v2f7d2dcfo89504c97666b6d7d@mail.gmail.com>
Message-ID: <447725.19101.qm@web34409.mail.mud.yahoo.com>

There is no particular special function called the
Green Function (sometimes named Green's Function). 
Green functions are special solutions of certain
partial differential equations (PDE), i.e. they are a
certain class of functions.  It's like saying
Orthogonal Polynomials -- that's a class, not a
particular type. 

Green functions have the property that when used in
the PDE they yield a Dirac delta function.  That makes
them useful for the solution of the same PDE, but with
a nonhomogeneous term added to the "other side."

So each PDE has it's own Green function.  Eg. the
Helmholtz equation,

 D^2 G + k^2 G=0  (D^2 = Laplacian)

has the Green function  G(r,r')= i H_0(k|r'-r|)/4,
where H_0 is a Hankel function and i=sqrt(-1).

I suggest you look at some texts (maybe mathematical
physics books) for the subject.  Even a Google search
will give you some information.

-- Lou Pecora

--- Natali Melgarejo Diaz <nmelgarejodiaz at gmail.com>
wrote:

> Hi to everyone i'm new in this list...i have a
> question about a
> formula i haven't found yet..., do you know the
> "Green function"??. Is
> there anything similar in Scipy??, 'cause in Matllab
> there is one but
> i haven't found it in Scipy yet.
> 
> Thanks in advance to everyone ;))
> 
> 
> ********Natali********



      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs


From fie.pye at atlas.cz  Tue Dec 11 14:03:00 2007
From: fie.pye at atlas.cz (Pye Fie)
Date: Tue, 11 Dec 2007 19:03:00 GMT
Subject: [SciPy-user] Building VTK
Message-ID: <b1c9768ec27c4f5d83c74ea10e047150@dd3fc183981f4e919f144122956c2934>

Hello.

I would like to build and install VTK and MayaVi2. 
PC: 2x Dual-core Opteron
275
OS: CentOS 5.0, kernel 2.6.18-8.1.15.el5
OS compiler: gcc version 4.2.0, gfortran

I
have already built and installed python2.5.1 and other module with compiler gcc3.4.6.
Now I can't build VTK. I am building with gcc3.4.6. Building stops on the following
error:

[ 74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkZLibDataCompressorTcl.o
[
74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkIOTCLInit.o
Linking CXX shared
library ../bin/libvtkIOTCL.so
[ 74%] Built target vtkIOTCL
Scanning dependencies
of target vtkParseOGLExt
[ 74%] Building CXX object Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/ParseOGLExt.o
[
74%] Building CXX object Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o
Linking
CXX executable ../../bin/vtkParseOGLExt
CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o:
In function `std::__verify_grouping(char const*, unsigned long, std::basic_string<char,
std::char_traits<char>, std::allocator<char> > const&)':
Tokenizer.cxx:(.text+0x19):
undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char>
>::size() const'
Tokenizer.cxx:(.text+0x70): undefined reference to `std::basic_string<char,
std::char_traits<char>, std::allocator<char> >::operator[](unsigned long) const'
Tokenizer.cxx:(.text+0xb0):
undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char>
>::operator[](unsigned long) const'
Tokenizer.cxx:(.text+0xdd): undefined reference
to `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::operator[](unsigned
long) const'
CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o: In function `Tokenizer::Tokenizer(char
const*, char const*)':
Tokenizer.cxx:(.text+0x12a): undefined reference to `std::allocator<char>::allocator()'
Tokenizer.cxx:(.text+0x13b):
undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char>
>::basic_string(char const*, std::allocator<char> const&)'
Tokenizer.cxx:(.text+0x14e):
undefined reference to `std::allocator<char>::~allocator()'
Tokenizer.cxx:(.text+0x164):
undefined reference to `std::allocator<char>::~allocator()'
Tokenizer.cxx:(.text+0x16d):
undefined reference to `std::allocator<char>::allocator()'
Tokenizer.cxx:(.text+0x182):
undefined reference to `std::basic_string<char, std::char_traits<char>, std::allocator<char>
>::basic_string(char const*, std::allocator<char> const&)'

The list continues. It
seams that there is something wrong with OpenGL but I can't find what is wrong. Could
anybody help me?
I have attached more detailed information about VTK configuration,
building output and building script.

Best regards
fie pye
------------------------------------------

http://search.atlas.cz/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: vtk_build.tar.gz
Type: application/x-gzip
Size: 30276 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071211/957c380c/attachment.bin>

From clovisgo at gmail.com  Tue Dec 11 14:09:34 2007
From: clovisgo at gmail.com (Clovis Goldemberg)
Date: Tue, 11 Dec 2007 17:09:34 -0200
Subject: [SciPy-user] Small problem with scipy.signal
Message-ID: <6f239f130712111109r139e0d47y2c5c2ab273892094@mail.gmail.com>

   I am having some problems with scipy.signal
   (specifically with scipy.signal.step) which can be
   explained by the following text,
   which was edited with some comments.
   Configuration is also given in the text.

   #####################################
   Microsoft Windows XP [Version 5.1.2600]
   (C) Copyright 1985-2001 Microsoft Corp.
   C:\>d:\python25\python
   Python 2.5.1 (r251:54863, Apr 18 2007,
   win32
   Type "help", "copyright", "credits" or
   >>> import scipy.signal as ss
   >>> import scipy
   >>> scipy.__version__
   '0.6.0'
   >>> import numpy
   >>> numpy.__version__
   '1.0.4'
   #####################################
   # numpy.test(10,2) runs without any problem
   # scipy.test(10,2) runs without any problem
   #####################################

- Ignored:
   >>> num=[1.0]
   >>> den=[1.0,1.0]
   >>> ss.step((num,den))
   #####################################
   # This works fine!
   # The transfer function would be "1/(1s+1)"
   # which is a simple low-pass filter.
   #####################################
   (array([ 0.  ,  0.07,  0.14,  0.21,  0.
           0.63,  0.7 ,  0.77,  0.84,  0.9
           1.26,  1.33,  1.4 ,  1.47,  1.5
           1.89,  1.96,  2.03,  2.1 ,  2.1
           2.52,  2.59,  2.66,  2.73,  2.8
           3.15,  3.22,  3.29,  3.36,  3.4
           3.78,  3.85,  3.92,  3.99,  4.0
           4.41,  4.48,  4.55,  4.62,  4.6
           5.04,  5.11,  5.18,  5.25,  5.3
           5.67,  5.74,  5.81,  5.88,  5.9
           6.3 ,  6.37,  6.44,  6.51,  6.5
   array([ 0.        ,  0.06760618,  0.1306
           0.29531191,  0.34295318,  0.387
           0.5034147 ,  0.53698693,  0.568
           0.65006225,  0.67372021,  0.695
           0.75340304,  0.77007451,  0.785
           0.82622606,  0.83797425,  0.848
           0.87754357,  0.88582238,  0.893
           0.91370641,  0.91954039,  0.924
           0.93918994,  0.94330107,  0.947
           0.95714787,  0.96004494,  0.962
           0.96980262,  0.97184415,  0.973
           0.97872026,  0.98015891,  0.981
           0.98500442,  0.98601822,  0.986
           0.9894328 ,  0.9901472 ,  0.990
           0.99255342,  0.99305685,  0.993
           0.99475248,  0.99510725,  0.995
           0.99630214,  0.99655213,  0.996
           0.99739416,  0.99757033,  0.997
           0.9981637 ,  0.99828784,  0.998
           0.99870598,  0.99879346,  0.998


   >>> den=[1.0,0.0]
   >>> ss.step((num,den))
   #####################################
   # This fails!
   # The transfer function would be "1/s"
   # which is a simple integrator.
   #####################################

   Traceback (most recent call last):
     File "<stdin>", line 1, in <module>
     File "d:\python25\lib\site-packages\s

       vals = lsim(sys, U, T, X0=X0)
     File "d:\python25\lib\site-packages\s

       xout[0] = X0
   IndexError: index out of bounds
   >>>

   #####################################
   Any suggestion?
   Thanks for your kind support.

- Done.



---------- Forwarded message ----------
From: "Clovis Goldemberg" <clovisgo at gmail.com>
To: scipy-user-request at scipy.org
Date: Mon, 10 Dec 2007 09:53:26 -0200
Subject: Small problem with "scipy.signal.step"
I'm having a small problem with scipy.signal, which is exposed in the
following text,
which was edited with some comments.

#####################################
Microsoft Windows XP [Version 5.1.2600]
(C) Copyright 1985-2001 Microsoft Corp.

C:\>d:\python25\python
Python 2.5.1 (r251:54863, Apr 18 2007,
win32
Type "help", "copyright", "credits" or
>>> import scipy.signal as ss
>>> import scipy
>>> scipy.__version__
'0.6.0'
>>> import numpy
>>> numpy.__version__
'1.0.4'

#####################################
# numpy.test(10,2) runs without any problem
# scipy.test(10,2) runs without any problem
#####################################

>>> num=[1.0]
>>> den=[1.0,1.0]
>>> ss.step((num,den))
#####################################
# This works fine!
# The transfer function would be "1/(1s+1)"
# which is a simple low-pass filter.
#####################################
(array([ 0.  ,  0.07,  0.14,  0.21,  0.
        0.63,  0.7 ,  0.77,  0.84,  0.9
        1.26,  1.33,  1.4 ,  1.47,  1.5
        1.89,  1.96,  2.03,  2.1 ,  2.1
        2.52,  2.59,  2.66,  2.73,  2.8
        3.15,  3.22,  3.29,  3.36,  3.4
        3.78,  3.85,  3.92,  3.99,  4.0
        4.41,  4.48,  4.55,  4.62,  4.6
        5.04,  5.11,  5.18,  5.25,  5.3
        5.67,  5.74,  5.81,  5.88,  5.9
        6.3 ,  6.37,  6.44,  6.51,  6.5
rray([ 0.        ,  0.06760618,  0.1306
        0.29531191,  0.34295318,  0.387
        0.5034147 ,  0.53698693,  0.568
        0.65006225,  0.67372021,  0.695
        0.75340304,  0.77007451,  0.785
        0.82622606,  0.83797425,  0.848
        0.87754357,  0.88582238,  0.893
        0.91370641,  0.91954039,  0.924
        0.93918994,  0.94330107 ,  0.947
        0.95714787,  0.96004494,  0.962
        0.96980262,  0.97184415,  0.973
        0.97872026,  0.98015891,  0.981
        0.98500442,  0.98601822,  0.986
        0.9894328 ,  0.9901472 ,  0.990
        0.99255342,  0.99305685,  0.993
        0.99475248,  0.99510725,  0.995
        0.99630214,  0.99655213,  0.996
        0.99739416,  0.99757033,  0.997
        0.9981637 ,  0.99828784,  0.998
        0.99870598,  0.99879346,  0.998


>>> den=[1.0,0.0]
>>> ss.step((num,den))
#####################################
# This fails!
# The transfer function would be "1/s"
# which is a simple integrator.
#####################################

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "d:\python25\lib\site-packages\s

    vals = lsim(sys, U, T, X0=X0)
  File "d:\python25\lib\site-packages\s

    xout[0] = X0
IndexError: index out of bounds
>>>

#####################################
Any suggestion?
Thanks for your kind support.
Clovis Goldemberg
Escola Politecnica da USP
Brasil


From matthieu.brucher at gmail.com  Tue Dec 11 14:56:05 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Tue, 11 Dec 2007 20:56:05 +0100
Subject: [SciPy-user] Building VTK
In-Reply-To: <b1c9768ec27c4f5d83c74ea10e047150@dd3fc183981f4e919f144122956c2934>
References: <b1c9768ec27c4f5d83c74ea10e047150@dd3fc183981f4e919f144122956c2934>
Message-ID: <e76aa17f0712111156x44ffc890h302484e7ea2f32b0@mail.gmail.com>

Hi,

This issue is related to a C++ problem in VTK, not Python, so we can't say
much. I advise you to post this on the VTK ML.

Matthieu

2007/12/11, Pye Fie <fie.pye at atlas.cz>:
>
> Hello.
>
> I would like to build and install VTK and MayaVi2.
> PC: 2x Dual-core Opteron
> 275
> OS: CentOS 5.0, kernel 2.6.18-8.1.15.el5
> OS compiler: gcc version 4.2.0, gfortran
>
> I
> have already built and installed python2.5.1 and other module with
> compiler gcc3.4.6.
> Now I can't build VTK. I am building with gcc3.4.6. Building stops on the
> following
> error:
>
> [ 74%] Building CXX object
> IO/CMakeFiles/vtkIOTCL.dir/vtkZLibDataCompressorTcl.o
> [
> 74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkIOTCLInit.o
> Linking CXX shared
> library ../bin/libvtkIOTCL.so
> [ 74%] Built target vtkIOTCL
> Scanning dependencies
> of target vtkParseOGLExt
> [ 74%] Building CXX object
> Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/ParseOGLExt.o
> [
> 74%] Building CXX object
> Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o
> Linking
> CXX executable ../../bin/vtkParseOGLExt
> CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o:
> In function `std::__verify_grouping(char const*, unsigned long,
> std::basic_string<char,
> std::char_traits<char>, std::allocator<char> > const&)':
> Tokenizer.cxx:(.text+0x19):
> undefined reference to `std::basic_string<char, std::char_traits<char>,
> std::allocator<char>
> >::size() const'
> Tokenizer.cxx:(.text+0x70): undefined reference to
> `std::basic_string<char,
> std::char_traits<char>, std::allocator<char> >::operator[](unsigned long)
> const'
> Tokenizer.cxx:(.text+0xb0):
> undefined reference to `std::basic_string<char, std::char_traits<char>,
> std::allocator<char>
> >::operator[](unsigned long) const'
> Tokenizer.cxx:(.text+0xdd): undefined reference
> to `std::basic_string<char, std::char_traits<char>, std::allocator<char>
> >::operator[](unsigned
> long) const'
> CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o: In function
> `Tokenizer::Tokenizer(char
> const*, char const*)':
> Tokenizer.cxx:(.text+0x12a): undefined reference to
> `std::allocator<char>::allocator()'
> Tokenizer.cxx:(.text+0x13b):
> undefined reference to `std::basic_string<char, std::char_traits<char>,
> std::allocator<char>
> >::basic_string(char const*, std::allocator<char> const&)'
> Tokenizer.cxx:(.text+0x14e):
> undefined reference to `std::allocator<char>::~allocator()'
> Tokenizer.cxx:(.text+0x164):
> undefined reference to `std::allocator<char>::~allocator()'
> Tokenizer.cxx:(.text+0x16d):
> undefined reference to `std::allocator<char>::allocator()'
> Tokenizer.cxx:(.text+0x182):
> undefined reference to `std::basic_string<char, std::char_traits<char>,
> std::allocator<char>
> >::basic_string(char const*, std::allocator<char> const&)'
>
> The list continues. It
> seams that there is something wrong with OpenGL but I can't find what is
> wrong. Could
> anybody help me?
> I have attached more detailed information about VTK configuration,
> building output and building script.
>
> Best regards
> fie pye
> ------------------------------------------
>
> http://search.atlas.cz/
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>


-- 
French PhD student
Website : http://matthieu-brucher.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/20071211/b3dd22e0/attachment.html>

From matthieu.brucher at gmail.com  Tue Dec 11 14:57:49 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Tue, 11 Dec 2007 20:57:49 +0100
Subject: [SciPy-user] Building VTK
In-Reply-To: <e76aa17f0712111156x44ffc890h302484e7ea2f32b0@mail.gmail.com>
References: <b1c9768ec27c4f5d83c74ea10e047150@dd3fc183981f4e919f144122956c2934>
	<e76aa17f0712111156x44ffc890h302484e7ea2f32b0@mail.gmail.com>
Message-ID: <e76aa17f0712111157w3aee2ef5n89165f3e55d6bfc8@mail.gmail.com>

Besides, you tell us you use GCC 3.4.6 but in your description, it is GCC
4.2...
The issue may be solved by linking against GCC libraries, but again, the VTK
ML may be a better place for the solution.

Matthieu

2007/12/11, Matthieu Brucher <matthieu.brucher at gmail.com>:
>
> Hi,
>
> This issue is related to a C++ problem in VTK, not Python, so we can't say
> much. I advise you to post this on the VTK ML.
>
> Matthieu
>
> 2007/12/11, Pye Fie < fie.pye at atlas.cz>:
> >
> > Hello.
> >
> > I would like to build and install VTK and MayaVi2.
> > PC: 2x Dual-core Opteron
> > 275
> > OS: CentOS 5.0, kernel 2.6.18-8.1.15.el5
> > OS compiler: gcc version 4.2.0, gfortran
> >
> > I
> > have already built and installed python2.5.1 and other module with
> > compiler gcc3.4.6.
> > Now I can't build VTK. I am building with gcc3.4.6. Building stops on
> > the following
> > error:
> >
> > [ 74%] Building CXX object
> > IO/CMakeFiles/vtkIOTCL.dir/vtkZLibDataCompressorTcl.o
> > [
> > 74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkIOTCLInit.o
> > Linking CXX shared
> > library ../bin/libvtkIOTCL.so
> > [ 74%] Built target vtkIOTCL
> > Scanning dependencies
> > of target vtkParseOGLExt
> > [ 74%] Building CXX object
> > Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/ParseOGLExt.o
> > [
> > 74%] Building CXX object
> > Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o
> > Linking
> > CXX executable ../../bin/vtkParseOGLExt
> > CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o:
> > In function `std::__verify_grouping(char const*, unsigned long,
> > std::basic_string<char,
> > std::char_traits<char>, std::allocator<char> > const&)':
> > Tokenizer.cxx:(.text+0x19):
> > undefined reference to `std::basic_string<char, std::char_traits<char>,
> > std::allocator<char>
> > >::size() const'
> > Tokenizer.cxx:(.text+0x70): undefined reference to
> > `std::basic_string<char,
> > std::char_traits<char>, std::allocator<char> >::operator[](unsigned
> > long) const'
> > Tokenizer.cxx: (.text+0xb0):
> > undefined reference to `std::basic_string<char, std::char_traits<char>,
> > std::allocator<char>
> > >::operator[](unsigned long) const'
> > Tokenizer.cxx:(.text+0xdd): undefined reference
> > to `std::basic_string<char, std::char_traits<char>, std::allocator<char>
> > >::operator[](unsigned
> > long) const'
> > CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o: In function
> > `Tokenizer::Tokenizer(char
> > const*, char const*)':
> > Tokenizer.cxx:(.text+0x12a): undefined reference to
> > `std::allocator<char>::allocator()'
> > Tokenizer.cxx:(.text+0x13b):
> > undefined reference to `std::basic_string<char, std::char_traits<char>,
> > std::allocator<char>
> > >::basic_string(char const*, std::allocator<char> const&)'
> > Tokenizer.cxx:(.text+0x14e):
> > undefined reference to `std::allocator<char>::~allocator()'
> > Tokenizer.cxx:(.text+0x164):
> > undefined reference to `std::allocator<char>::~allocator()'
> > Tokenizer.cxx:(.text+0x16d):
> > undefined reference to `std::allocator<char>::allocator()'
> > Tokenizer.cxx:(.text+0x182):
> > undefined reference to `std::basic_string<char, std::char_traits<char>,
> > std::allocator<char>
> > >::basic_string(char const*, std::allocator<char> const&)'
> >
> > The list continues. It
> > seams that there is something wrong with OpenGL but I can't find what is
> > wrong. Could
> > anybody help me?
> > I have attached more detailed information about VTK configuration,
> > building output and building script.
> >
> > Best regards
> > fie pye
> > ------------------------------------------
> >
> > http://search.atlas.cz/
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
> >
> >
>
>
> --
> French PhD student
> Website : http://matthieu-brucher.developpez.com/
> Blogs : http://matt.eifelle.com and http://blog.developpez.com/?blog=92
> LinkedIn : http://www.linkedin.com/in/matthieubrucher




-- 
French PhD student
Website : http://matthieu-brucher.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/20071211/81a816f8/attachment.html>

From hasslerjc at comcast.net  Tue Dec 11 15:01:24 2007
From: hasslerjc at comcast.net (John Hassler)
Date: Tue, 11 Dec 2007 15:01:24 -0500
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>	<475EAEC1.4020501@enthought.com>	<85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>
	<e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>
Message-ID: <475EEC94.3010609@comcast.net>

An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071211/ca9932cb/attachment.html>

From fperez.net at gmail.com  Tue Dec 11 15:05:04 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Tue, 11 Dec 2007 13:05:04 -0700
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>
	<475EAEC1.4020501@enthought.com>
	<85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>
	<e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>
Message-ID: <db6b5ecc0712111205s1f78dce2tfdf3c501db754d78@mail.gmail.com>

On Dec 11, 2007 9:45 AM, Matthieu Brucher <matthieu.brucher at gmail.com> wrote:
> You can't expect scipy to find the exact coefficients. sage is a symbolic
> package, it will find the correct answers, but scipy will find only an
> approximated one, up to the machine precision. This is what you see in your
> exemple.
> If you have integers, you could expect scipy to return long integers (exact
> result), but this is not the case as almost everything is converted into a
> float array before the actual C or Fortran routine is run.

In this case Sage isn't using anything symbolic though: the issue is
that Sage has from the ground up defined a type system where it knows
what field its inputs are being computed over.  So if a matrix is made
up of only integers, it knows that computations are to be performed in
exact arithmetic over Z, and does so accordingly.

Sage's origins are actually number theory, not symbolic computing, and
its strongest area is precisely in exact arithmetic, with enormous
sophistication for some of its algorithms.  Initially its (free)
symbolic capabilities were all obtained via calls to Maxima, though
now that Ondrej is actively helping with SymPy integration into Sage,
that is changing and SymPy is natively available as well, which means
that over time there will be more and more native (python) symbolic
support as well.

I'd agree though that for this kind of calculation, Sage is the tool to use.

Cheers,

f


From lorenzo.isella at gmail.com  Tue Dec 11 15:39:18 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Tue, 11 Dec 2007 21:39:18 +0100
Subject: [SciPy-user] Operations on Arrays in SciPy
Message-ID: <475EF576.8010503@gmail.com>

Dear All,
A  probably quick question:  SciPy's arrays are objects  with methods, 
are they not?
For instance,  consider the following:
import scipy as s
z=s.arange(30)
mean_z=s.mean(z)
mean_z_2=z.mean()

Is there a reason why one way of operating on the array should be 
preferable?
Is it the same thing to work out the mean of z as mean_z or mean_z_2?
Many thanks

Lorenzo


From tjhnson at gmail.com  Tue Dec 11 15:48:01 2007
From: tjhnson at gmail.com (Tom Johnson)
Date: Tue, 11 Dec 2007 12:48:01 -0800
Subject: [SciPy-user] maskedarray
Message-ID: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>

What is the status regarding 'maskedarray'?  When will this become
part (replace ma) of the standard distribution?  Also, what is the
recommended way to import sandbox modules so that code changes are
minimized when (if) the package becomes part of the standard
distribution.


From robert.kern at gmail.com  Tue Dec 11 16:19:32 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 11 Dec 2007 15:19:32 -0600
Subject: [SciPy-user] Operations on Arrays in SciPy
In-Reply-To: <475EF576.8010503@gmail.com>
References: <475EF576.8010503@gmail.com>
Message-ID: <475EFEE4.2060407@gmail.com>

Lorenzo Isella wrote:
> Dear All,
> A  probably quick question:  SciPy's arrays are objects  with methods, 
> are they not?

Yes.

> For instance,  consider the following:
> import scipy as s
> z=s.arange(30)
> mean_z=s.mean(z)
> mean_z_2=z.mean()
> 
> Is there a reason why one way of operating on the array should be 
> preferable?

scipy.mean() can accept anything that can be converted to an array, like a list
of numbers.

> Is it the same thing to work out the mean of z as mean_z or mean_z_2?

Yes.

-- 
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 stefan at sun.ac.za  Tue Dec 11 17:41:17 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Wed, 12 Dec 2007 00:41:17 +0200
Subject: [SciPy-user] Small problem with scipy.signal
In-Reply-To: <6f239f130712111109r139e0d47y2c5c2ab273892094@mail.gmail.com>
References: <6f239f130712111109r139e0d47y2c5c2ab273892094@mail.gmail.com>
Message-ID: <20071211224117.GG11744@mentat.za.net>

Hi Clovis

On Tue, Dec 11, 2007 at 05:09:34PM -0200, Clovis Goldemberg wrote:
>    I am having some problems with scipy.signal
>    (specifically with scipy.signal.step) which can be
>    explained by the following text,
>    which was edited with some comments.
>    Configuration is also given in the text.

[...]

>    >>> num=[1.0]
>    >>> den=[1.0,0.0]
>    >>> ss.step((num,den))
>    #####################################
>    # This fails!
>    # The transfer function would be "1/s"
>    # which is a simple integrator.
>    #####################################

Looks like the system is converted to state space.  After that,
somewhere along the line A is inverted, which causes the problem.  I
think a person can get away with only inverting [I-A], but I'll have
to poke around the source a bit to be sure.

Regards
St?fan


From roger.herikstad at gmail.com  Tue Dec 11 21:28:12 2007
From: roger.herikstad at gmail.com (Roger Herikstad)
Date: Wed, 12 Dec 2007 10:28:12 +0800
Subject: [SciPy-user] Correlation coefficients
Message-ID: <bae941750712111828h6d64e34x70a062379b7df269@mail.gmail.com>

Hi all, I was wondering if there is an efficient way of determining the
degree of correlation between an array and and pool of reference arrays (>
10,000). I'm using corrcoef when the pool is small ( <1,000), but I have
cases where I need to compare against more than 10,000 arrays, in which case
corrcoef is slow. Any suggestions? Thanks!

~ Roger
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071212/7102a352/attachment.html>

From bnsulman at wisc.edu  Wed Dec 12 01:04:26 2007
From: bnsulman at wisc.edu (Ben Sulman)
Date: Wed, 12 Dec 2007 06:04:26 +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.20071212T055709-589@post.gmane.org>

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

> 
> > > 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.
> 



Hi, I was just looking through this list for ideas on how I could better use
scipy.  My issue is that I'm working with time series data, with missing values,
that is in half-hour increments.  As far as I can tell, with the current
implementation of timeseries many operations will only work if the data has a
point for every increment of a built in frequency.  This means I have to either
put my data in hourly form, which would require averaging, or convert it to
minute frequency with missing values, which expands the size of my data by 30x.
 Is there a way around this, where I could just process half-hourly data?

Thanks for your help!  Timeseries looks like a really useful package if I can
get it to play well with my data!

Ben Sulman








From cscheit at lstm.uni-erlangen.de  Wed Dec 12 05:57:14 2007
From: cscheit at lstm.uni-erlangen.de (Christoph Scheit)
Date: Wed, 12 Dec 2007 11:57:14 +0100
Subject: [SciPy-user] read ascii data
In-Reply-To: <mailman.10357.1197403323.2301.scipy-user@scipy.org>
References: <mailman.10357.1197403323.2301.scipy-user@scipy.org>
Message-ID: <200712121157.14457.cscheit@lstm.uni-erlangen.de>

Hi All,

I have a file containing ascii data related to a block-structured grid.
I could use readtxt, but I would like to have the possibility to read only
a certain range of rows out of the file. Lets say I have ten blocks in my grid
and each block 10 rows of data.
So in one read I would like to read for instance line 30 to line 39
I can use skip rows, but how can I limit the loadtxt to the upper limit?
Is there something already available or do I have to do it on my own?

Thanks in advance,

Christoph


From xavier.barthelemy at cmla.ens-cachan.fr  Wed Dec 12 07:57:32 2007
From: xavier.barthelemy at cmla.ens-cachan.fr (Xavier Barthelemy)
Date: Wed, 12 Dec 2007 13:57:32 +0100
Subject: [SciPy-user] read ascii data
In-Reply-To: <200712121157.14457.cscheit@lstm.uni-erlangen.de>
References: <mailman.10357.1197403323.2301.scipy-user@scipy.org>
	<200712121157.14457.cscheit@lstm.uni-erlangen.de>
Message-ID: <475FDABC.7000408@cmla.ens-cachan.fr>

hi

in scipy, or numpy I can't recall, there's a function name io.read_array

data= 
io.read_array(filename,columns=(0,-1),lines=(0,-1),atype='thetypeyouwant')

columns and lines can be all like int his example, or you can just 
choose where to begin and where to end, or just a discrete list.

check the Doc, it is relatively well explained

Cheers
Xavier

Christoph Scheit a ?crit :
> Hi All,
> 
> I have a file containing ascii data related to a block-structured grid.
> I could use readtxt, but I would like to have the possibility to read only
> a certain range of rows out of the file. Lets say I have ten blocks in my grid
> and each block 10 rows of data.
> So in one read I would like to read for instance line 30 to line 39
> I can use skip rows, but how can I limit the loadtxt to the upper limit?
> Is there something already available or do I have to do it on my own?
> 
> Thanks in advance,
> 
> Christoph
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user




From odonnems at yahoo.com  Wed Dec 12 08:55:49 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Wed, 12 Dec 2007 05:55:49 -0800 (PST)
Subject: [SciPy-user] read ascii data
Message-ID: <255482.8679.qm@web58012.mail.re3.yahoo.com>

I am a new python/scipy user, but my first question is what type of grid data are you working with. I am not sure how you are defining 'grid' in other words. If you are working with spatial raster data, there is a library called GDAL. If the data has no spatial content you may try PIL library. 

I have also worked with reading comma delimited ascii files and you could also enumerate and skip all lines except for those you are interested in pretty easily and quickly. 

These are a few ideas, but which method you should select likely depends on the format of the data as well as what you plan on doing with the data that you are reading.

I hope this gives you some ideas,
Michael



----- Original Message ----
From: Christoph Scheit <cscheit at lstm.uni-erlangen.de>
To: scipy-user at scipy.org
Sent: Wednesday, December 12, 2007 3:57:14 AM
Subject: [SciPy-user] read ascii data


Hi All,

I have a file containing ascii data related to a block-structured grid.
I could use readtxt, but I would like to have the possibility to read
 only
a certain range of rows out of the file. Lets say I have ten blocks in
 my grid
and each block 10 rows of data.
So in one read I would like to read for instance line 30 to line 39
I can use skip rows, but how can I limit the loadtxt to the upper
 limit?
Is there something already available or do I have to do it on my own?

Thanks in advance,

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






      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071212/c1f72495/attachment.html>

From pgmdevlist at gmail.com  Wed Dec 12 09:24:49 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Wed, 12 Dec 2007 09:24:49 -0500
Subject: [SciPy-user] maskedarray
In-Reply-To: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
Message-ID: <200712120924.49482.pgmdevlist@gmail.com>

On Tuesday 11 December 2007 15:48:01 Tom Johnson wrote:
> What is the status regarding 'maskedarray'?  When will this become
> part (replace ma) of the standard distribution?  Also, what is the
> recommended way to import sandbox modules so that code changes are
> minimized when (if) the package becomes part of the standard
> distribution.


Mmh, no answer from our leaders ? Well, for now maskedarray lives in the scipy 
sandbox, but it probably could be transferred to the numpy sandbox. There's 
still no particular decision I'm aware of about putting it the official 
release. More feedback is probably needed from users.


From bsouthey at gmail.com  Wed Dec 12 09:42:12 2007
From: bsouthey at gmail.com (Bruce Southey)
Date: Wed, 12 Dec 2007 08:42:12 -0600
Subject: [SciPy-user] Correlation coefficients
In-Reply-To: <bae941750712111828h6d64e34x70a062379b7df269@mail.gmail.com>
References: <bae941750712111828h6d64e34x70a062379b7df269@mail.gmail.com>
Message-ID: <bbcd77d00712120642w666dee25y2f40915d3b37feb1@mail.gmail.com>

Hi,
What exactly do you want as an outcome? In particular, is there a
simpler outcome such as only the covariance?

Are you correlating one array with each array of the pool (or against
the pool average)?
Or correlating each array in the pool with each other?

You may only need the covariance or crossproduct if the variance of
each array are essentially the same or covariance is sufficient. In
that case a array by array multiplication (pool as a multi-dimensional
array) will be sufficient. If not, create your own correlation
function where you store the mean and variance of each array the pool
so you only compute it once.


Regards
Bruce






On Dec 11, 2007 8:28 PM, Roger Herikstad <roger.herikstad at gmail.com> wrote:
> Hi all,
>  I was wondering if there is an efficient way of determining the degree of
> correlation between an array and and pool of reference arrays (> 10,000).
> I'm using corrcoef when the pool is small ( <1,000), but I have cases where
> I need to compare against more than 10,000 arrays, in which case corrcoef is
> slow. Any suggestions? Thanks!
>
> ~ Roger
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


From odonnems at yahoo.com  Wed Dec 12 12:02:45 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Wed, 12 Dec 2007 09:02:45 -0800 (PST)
Subject: [SciPy-user] using STL in scipy weave inline
Message-ID: <92324.71222.qm@web58005.mail.re3.yahoo.com>

Vincent and Eric thank you for the information. Sorry I took a while to get back to this, but I was side tracked while building another library that I needed for this program. I do have a couple more questions about headers and includes.

I tried to use #include <map> at the beginning of the script originally but when I did this I got a bunch of errors about the <map>. My question: is there a way to list the headers at the beginning of the script rather than passing through the weave.inline code? I am interested in knowing this because I will need to do this for several things and iI would like to minimize the information  in the command line as much as possible.

Secondly, can someone explain the py:: code. I have seen this but I got an error when I try to include it in my code? I am not getting an error with the code Eric attached.

Third, can someone recommend any documentation or books that may help me with the interface between python and c++? For example, understanding the py::. I have not seen a whole lot on line.

Thanks again for the help and the code!!
Michael


----- Original Message ----
From: eric jones <eric at enthought.com>
To: SciPy Users List <scipy-user at scipy.org>
Sent: Monday, December 10, 2007 4:17:06 PM
Subject: Re: [SciPy-user] using STL in scipy weave inline


Michael ODonnell wrote:
> OS Microsoft XP SP2 x86
> python 2.5.1
> compiler: mingw
>
> Does anyone have experience using multimap or the standard template 
> library within weave inline? I need to create a map that contains 
> numeric keys which are paired to a string, but I get an error when 
> declaring the multimap function. If i include <map> I get other
 errors 
> about namespace and if don't then I get an error for not declaring
 the 
> multimap. Neither method works.
There is no problem using the stl from weave.  In fact, it uses 
std::string internally.

You can use

    headers=["<map>"]

as a keyword to the inline call to get map into the namespace.  Once
 you 
do that, you'll need to either declare "using namespace std;" at the
 top 
of your code or use std::multimap and std::string to refer to the 
classes in the stl.  Once you've done that, you'll still have some 
issues as multimap doesn't support the operator [], and your C++
 strings 
have single quotes around them instead of double quotes.

I've attached an example for your code that instead works using a 
standard python dictionary (use py::dict within weave).  The output is 
shown here:

In [33]: run we_ex.py
<weave: compiling>

In [34]: res
Out[34]:
{0.050000000000000003: 'fESE',
 0.10000000000000001: 'fSSE',
 0.14999999999999999: 'fFlat',
 0.25: 'fNNE',
 0.40000000000000002: 'fSSW',
 0.5: 'fWNW',
 0.55000000000000004: 'fNNW',
 1.1000000000000001: 'fND',
 1.5: 'fWSW'}

If you need to use the stl for some other reason, then refer to this to
 
see how to do it: http://www.sgi.com/tech/stl/Multimap.html

Good luck,
eric
>
> TIA,
> Michael
>
> Example of code:
>
> total = (winx * winy * 1.0) #Total number of cells to evaluate in
 kernel
>
> code = r"""
>             double ND, NNW, WNW, WSW,SSW, SSE, ESE, ENE, NNE, Flat;
>             ND = 22.0;
>             NNW = 11.0;
>             WNW = 10.0;
>             WSW = 30.0;
>             SSW = 8.0;
>             SSE = 2.0;
>             ESE = 1.0;
>             ENE = 5.0;
>             NNE = 5.0;
>             Flat = 3.0;
>
>             //Calculate the frequency of the counts
>             fND = (ND / total);   
>             fNNW = (NNW / total);  
>             fWNW = (WNW / total); 
>             fWSW = (WSW / total); 
>             fSSW = (SSW / total);  
>             fSSE = (SSE / total);  
>             fESE = (ESE / total); 
>             fENE = (ENE / total); 
>             fNNE = (NNE / total);  
>             fFlat = (Flat / total); 
>        
>             //Create a map for pairs: frequency, string
>             multimap<double, string> map_freq; //Create empty map
>        
>             map_freq[fND) = 'fND';
>             map_freq[fNNW] = 'fNNW';
>             map_freq[fWNW] = 'fWNW';
>             map_freq[fWSW] = 'fWSW';
>             map_freq[fSSW] = 'fSSW';
>             map_freq[fSSE] = 'fSSE';
>             map_freq[fESE] = 'fESE';
>             map_freq[fENE] = 'fENE';
>             map_freq[fNNE] = 'fNNE';
>             map_freq[fFlat] = 'fFlat';
>             """
>
> weave.inline(code,['total'], type_converters=converters.blitz, 
> compiler='gcc')
>
>
 ------------------------------------------------------------------------
> Be a better friend, newshound, and know-it-all with Yahoo! Mobile.
 Try 
> it now. 
>
 <http://us.rd.yahoo.com/evt=51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ%20> 
>
>
 ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



-----Inline Attachment Follows-----

from scipy import weave
from scipy.weave import converters


total = 20.0 # (winx * winy * 1.0) #Total number of cells to evaluate
 in kernel

code = r"""
            double ND, NNW, WNW, WSW,SSW, SSE, ESE, ENE, NNE, Flat;
            ND = 22.0;
            NNW = 11.0;
            WNW = 10.0;
            WSW = 30.0;
            SSW = 8.0;
            SSE = 2.0;
            ESE = 1.0;
            ENE = 5.0;
            NNE = 5.0;
            Flat = 3.0;

            //Calculate the frequency of the counts
            double fND, fNNW, fWNW, fWSW,fSSW, fSSE, fESE, fENE, fNNE,
 fFlat;
            fND = (ND / total);   
            fNNW = (NNW / total);  
            fWNW = (WNW / total); 
            fWSW = (WSW / total); 
            fSSW = (SSW / total);  
            fSSE = (SSE / total);  
            fESE = (ESE / total); 
            fENE = (ENE / total); 
            fNNE = (NNE / total);  
            fFlat = (Flat / total); 
       
            //Create a map for pairs: frequency, string
            //std::multimap<double, std::string> map_freq; //Create
 empty map
            py::dict map_freq;
       
            map_freq[fND] = "fND";
            map_freq[fNNW] = "fNNW";
            map_freq[fWNW] = "fWNW";
            map_freq[fWSW] = "fWSW";
            map_freq[fSSW] = "fSSW";
            map_freq[fSSE] = "fSSE";
            map_freq[fESE] = "fESE";
            map_freq[fENE] = "fENE";
            map_freq[fNNE] = "fNNE";
            map_freq[fFlat] = "fFlat";
            
            return_val = map_freq;
            """

res = weave.inline(code,['total'], #headers= ['<map>'],
                   compiler='gcc')






      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071212/baf6d42c/attachment.html>

From matthew.brett at gmail.com  Wed Dec 12 12:42:17 2007
From: matthew.brett at gmail.com (Matthew Brett)
Date: Wed, 12 Dec 2007 17:42:17 +0000
Subject: [SciPy-user] maskedarray
In-Reply-To: <200712120924.49482.pgmdevlist@gmail.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<200712120924.49482.pgmdevlist@gmail.com>
Message-ID: <1e2af89e0712120942t65a88f21xf5d83aa1f9966d9a@mail.gmail.com>

Hi,

> Mmh, no answer from our leaders ? Well, for now maskedarray lives in the scipy
> sandbox, but it probably could be transferred to the numpy sandbox. There's
> still no particular decision I'm aware of about putting it the official
> release. More feedback is probably needed from users.

Perhaps by making the shift in SVN? (!)

Matthew


From matthieu.brucher at gmail.com  Wed Dec 12 13:01:14 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Wed, 12 Dec 2007 19:01:14 +0100
Subject: [SciPy-user] using STL in scipy weave inline
In-Reply-To: <92324.71222.qm@web58005.mail.re3.yahoo.com>
References: <92324.71222.qm@web58005.mail.re3.yahoo.com>
Message-ID: <e76aa17f0712121001j4d8cd03dm2b1b01cfa14bbe0a@mail.gmail.com>

2007/12/12, Michael ODonnell <odonnems at yahoo.com>:
>
> I tried to use #include <map> at the beginning of the script originally
> but when I did this I got a bunch of errors about the <map>. My question: is
> there a way to list the headers at the beginning of the script rather than
> passing through the weave.inline code? I am interested in knowing this
> because I will need to do this for several things and iI would like to
> minimize the information  in the command line as much as possible.
>


You can't include it in your code directly as the code is in fact put in
another function. The cleanest way is to use the headers keyword.


Secondly, can someone explain the py:: code. I have seen this but I got an
> error when I try to include it in my code? I am not getting an error with
> the code Eric attached.
>


py:: est the namespace where the wrappers between Python and C++ objects are
made. Weave seems to use the SCXX package according to the source code (but
I'm not sure it is the real package).


Third, can someone recommend any documentation or books that may help me
> with the interface between python and c++? For example, understanding the
> py::. I have not seen a whole lot on line.
>

There are no such book at the moment. There are tutorials for SWIG, less for
Boost.Python, but for weave, I don't know.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071212/d61ed04d/attachment.html>

From fperez.net at gmail.com  Wed Dec 12 14:17:30 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Wed, 12 Dec 2007 12:17:30 -0700
Subject: [SciPy-user] maskedarray
In-Reply-To: <200712120924.49482.pgmdevlist@gmail.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<200712120924.49482.pgmdevlist@gmail.com>
Message-ID: <db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>

On Dec 12, 2007 7:24 AM, Pierre GM <pgmdevlist at gmail.com> wrote:
> On Tuesday 11 December 2007 15:48:01 Tom Johnson wrote:
> > What is the status regarding 'maskedarray'?  When will this become
> > part (replace ma) of the standard distribution?  Also, what is the
> > recommended way to import sandbox modules so that code changes are
> > minimized when (if) the package becomes part of the standard
> > distribution.
>
>
> Mmh, no answer from our leaders ? Well, for now maskedarray lives in the scipy
> sandbox, but it probably could be transferred to the numpy sandbox. There's
> still no particular decision I'm aware of about putting it the official
> release. More feedback is probably needed from users.

Well, I'm certainly not one of the leaders, but it's probably worth
mentioning that this weekend Jarrod is organizing a small sprint at UC
Berkeley to try to flush out a number of things for numpy/scipy.  And
last weekend John Hunter and I taught a workshop at NCAR here in
Boulder, where they expressed a LOT of interest in seeing the MA
situation resolved, and were willing to commit resources to making it
happen (their codes rely on MA throughout).  So I added MA to the
sprint todo list already, counting on NCAR's resources once a decision
is made.  I'd meant to mention it here and just hadn't had the time
for it.

So it would be great if perhaps we could be in touch over the weekend
(Fri/Sat/Sun) to touch bases on this at some point, to get your
feedback.  If you have skype access, perhaps you could email me
directly and we can then set up a time(s) when a brief skype
conversation might be feasible.

Cheers,

f


From stefan at sun.ac.za  Wed Dec 12 15:02:37 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Wed, 12 Dec 2007 22:02:37 +0200
Subject: [SciPy-user] maskedarray
In-Reply-To: <db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<200712120924.49482.pgmdevlist@gmail.com>
	<db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
Message-ID: <20071212200237.GB14894@mentat.za.net>

On Wed, Dec 12, 2007 at 12:17:30PM -0700, Fernando Perez wrote:
> On Dec 12, 2007 7:24 AM, Pierre GM <pgmdevlist at gmail.com> wrote:
> > On Tuesday 11 December 2007 15:48:01 Tom Johnson wrote:
> > > What is the status regarding 'maskedarray'?  When will this become
> > > part (replace ma) of the standard distribution?  Also, what is the
> > > recommended way to import sandbox modules so that code changes are
> > > minimized when (if) the package becomes part of the standard
> > > distribution.
> >
> >
> > Mmh, no answer from our leaders ? Well, for now maskedarray lives in the scipy
> > sandbox, but it probably could be transferred to the numpy sandbox. There's
> > still no particular decision I'm aware of about putting it the official
> > release. More feedback is probably needed from users.
> 
> Well, I'm certainly not one of the leaders, but it's probably worth
> mentioning that this weekend Jarrod is organizing a small sprint at UC
> Berkeley to try to flush out a number of things for numpy/scipy.  And
> last weekend John Hunter and I taught a workshop at NCAR here in
> Boulder, where they expressed a LOT of interest in seeing the MA
> situation resolved, and were willing to commit resources to making it
> happen (their codes rely on MA throughout).  So I added MA to the
> sprint todo list already, counting on NCAR's resources once a decision
> is made.  I'd meant to mention it here and just hadn't had the time
> for it.

If you guys organise sprints, please let the rest of us know so that
we can work together online (if possible).

Thanks!
St?fan


From bsouthey at gmail.com  Wed Dec 12 16:06:25 2007
From: bsouthey at gmail.com (Bruce Southey)
Date: Wed, 12 Dec 2007 15:06:25 -0600
Subject: [SciPy-user] maskedarray
In-Reply-To: <db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<200712120924.49482.pgmdevlist@gmail.com>
	<db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
Message-ID: <bbcd77d00712121306q12c30f38uaf957065b084d638@mail.gmail.com>

Hi,
This has been raised on this list in the past. IIRC the main issue is
one of sufficient testing particularly by users. Of course this is a
little like what comes first as people probably just use ma as it is
part of numpy.

I would like to see it in but I don't use it heavily to sufficiently
comment on it.

Regards
Bruce



On Dec 12, 2007 1:17 PM, Fernando Perez <fperez.net at gmail.com> wrote:
> On Dec 12, 2007 7:24 AM, Pierre GM <pgmdevlist at gmail.com> wrote:
> > On Tuesday 11 December 2007 15:48:01 Tom Johnson wrote:
> > > What is the status regarding 'maskedarray'?  When will this become
> > > part (replace ma) of the standard distribution?  Also, what is the
> > > recommended way to import sandbox modules so that code changes are
> > > minimized when (if) the package becomes part of the standard
> > > distribution.
> >
> >
> > Mmh, no answer from our leaders ? Well, for now maskedarray lives in the scipy
> > sandbox, but it probably could be transferred to the numpy sandbox. There's
> > still no particular decision I'm aware of about putting it the official
> > release. More feedback is probably needed from users.
>
> Well, I'm certainly not one of the leaders, but it's probably worth
> mentioning that this weekend Jarrod is organizing a small sprint at UC
> Berkeley to try to flush out a number of things for numpy/scipy.  And
> last weekend John Hunter and I taught a workshop at NCAR here in
> Boulder, where they expressed a LOT of interest in seeing the MA
> situation resolved, and were willing to commit resources to making it
> happen (their codes rely on MA throughout).  So I added MA to the
> sprint todo list already, counting on NCAR's resources once a decision
> is made.  I'd meant to mention it here and just hadn't had the time
> for it.
>
> So it would be great if perhaps we could be in touch over the weekend
> (Fri/Sat/Sun) to touch bases on this at some point, to get your
> feedback.  If you have skype access, perhaps you could email me
> directly and we can then set up a time(s) when a brief skype
> conversation might be feasible.
>
> Cheers,
>
> f
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From travis at enthought.com  Wed Dec 12 16:21:29 2007
From: travis at enthought.com (Travis Vaught)
Date: Wed, 12 Dec 2007 15:21:29 -0600
Subject: [SciPy-user] maskedarray
In-Reply-To: <db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<200712120924.49482.pgmdevlist@gmail.com>
	<db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
Message-ID: <C0C14B2C-2B65-41BB-AEE8-606C938D36C7@enthought.com>

On Dec 12, 2007, at 1:17 PM, Fernando Perez wrote:

>> ...
>
> Well, I'm certainly not one of the leaders, but it's probably worth
> mentioning that this weekend Jarrod is organizing a small sprint at UC
> Berkeley to try to flush out a number of things for numpy/scipy.  And
> last weekend John Hunter and I taught a workshop at NCAR here in
> Boulder, where they expressed a LOT of interest in seeing the MA
> situation resolved, and were willing to commit resources to making it
> happen (their codes rely on MA throughout).  So I added MA to the
> sprint todo list already, counting on NCAR's resources once a decision
> is made.  I'd meant to mention it here and just hadn't had the time
> for it.
>
> So it would be great if perhaps we could be in touch over the weekend
> (Fri/Sat/Sun) to touch bases on this at some point, to get your
> feedback.  If you have skype access, perhaps you could email me
> directly and we can then set up a time(s) when a brief skype
> conversation might be feasible.
>
> Cheers,
>
> f

Just to remind folks, there is a scipy channel at irc.freenode.net as  
well...that might be nice to allow more folks to participate, and to  
have an archive of the conversation.


From fie.pye at atlas.cz  Wed Dec 12 16:46:09 2007
From: fie.pye at atlas.cz (Pye Fie)
Date: Wed, 12 Dec 2007 21:46:09 GMT
Subject: [SciPy-user] Building VTK
Message-ID: <abcdd5af48e941979fa87eee4d6432fa@0d033be14cd945f3a6fde4b3608c857b>

 
Hi Metthieu.
Than you for your response. I posted my question on VTK ML first.
http://public.kitware.com/pipermail/vtkusers/2007-December/093627.html
Unfortunately I didn't get any answer until now. 

You asker about my compiler. 
CentOS 5.0 is compiled with gcc 4.1. I installed also gcc 4.2 and gcc 3.4.6 on the PC.

I would like to have a computational environment based on Python and its module. I compiled Python and module Ipython, SciPy, NumPy, tables, matplotlib, wxpython ......... and libraries such as BLAS, LAPACK, ATLAS, HDF5, ........ with gcc3.4.6. Now I continue with compiling VTK with gcc3.4.6 too. More information about compilation setup is the file CMakeCache.txt.

Best regards.
Fie pye


Od: Matthieu Brucher
P?ijato: 11.12.2007 21:18:04
P?edm?t: Re: [SciPy-user] Building VTK
Besides, you tell us you use GCC 3.4.6 but in your description, it is GCC 4.2...
The issue may be solved by linking against GCC libraries, but again, the VTK ML may be a better place for the solution.
Matthieu
2007/12/11, Matthieu Brucher <matthieu.brucher at gmail.com>: Hi,
This issue is related to a C++ problem in VTK, not Python, so we can't say much. I advise you to post this on the VTK ML.
Matthieu
2007/12/11, Pye Fie < fie.pye at atlas.cz>:Hello.
I would like to build and install VTK and MayaVi2. 
PC: 2x Dual-core Opteron
275
OS: CentOS 5.0, kernel 2.6.18-8.1.15.el5
OS compiler: gcc version 4.2.0, gfortran
I
have already built and installed python2.5.1 and other module with compiler gcc3.4.6 . 
Now I can't build VTK. I am building with gcc3.4.6. Building stops on the following
error:
[ 74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkZLibDataCompressorTcl.o
[
74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkIOTCLInit.o 
Linking CXX shared
library ../bin/libvtkIOTCL.so
[ 74%] Built target vtkIOTCL
Scanning dependencies
of target vtkParseOGLExt
[ 74%] Building CXX object Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/ParseOGLExt.o 
[
74%] Building CXX object Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o
Linking
CXX executable ../../bin/vtkParseOGLExt
CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o:
In function `std::__verify_grouping(char const*, unsigned long, std::basic_string
std::char_traits, std::allocator > const&amp;)':
Tokenizer.cxx:(.text+0x19):
undefined reference to `std::basic_string, std::allocator 
>::size() const'
Tokenizer.cxx:(.text+0x70): undefined reference to `std::basic_string
std::char_traits, std::allocator >::operator[](unsigned long) const'
Tokenizer.cxx : (.text+0xb0):
undefined reference to `std::basic_string, std::allocator
>::operator[](unsigned long) const'
Tokenizer.cxx:(.text+0xdd): undefined reference 
to `std::basic_string, std::allocator >::operator[](unsigned
long) const'
CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o: In function `Tokenizer::Tokenizer(char 
const*, char const*)':
Tokenizer.cxx:(.text+0x12a): undefined reference to `std::allocator::allocator()'
Tokenizer.cxx:(.text+0x13b):
undefined reference to `std::basic_string, std::allocator 
>::basic_string(char const*, std::allocator const&amp;)'
Tokenizer.cxx:(.text+0x14e):
undefined reference to `std::allocator::~allocator()'
Tokenizer.cxx:(.text+0x164):
undefined reference to `std::allocator::~allocator()' 
Tokenizer.cxx:(.text+0x16d):
undefined reference to `std::allocator::allocator()'
Tokenizer.cxx:(.text+0x182):
undefined reference to `std::basic_string, std::allocator 
>::basic_string(char const*, std::allocator const&amp;)'
The list continues. It
seams that there is something wrong with OpenGL but I can't find what is wrong. Could
anybody help me? 
I have attached more detailed information about VTK configuration,
building output and building script.
Best regards
fie pye
------------------------------------------
http://search.atlas.cz/
_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org
http://projects.scipy.org/mailman/listinfo/scipy-user
-- 
French PhD student
Website :  http://matthieu-brucher.developpez.com/ 
Blogs : http://matt.eifelle.com and  http://blog.developpez.com/?blog=92
LinkedIn :  http://www.linkedin.com/in/matthieubrucher 
-- 
French PhD student
Website : http://matthieu-brucher.developpez.com/
Blogs :  http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn : http://www.linkedin.com/in/matthieubrucher  
------------------------------------------

http://search.atlas.cz/

From fperez.net at gmail.com  Wed Dec 12 17:23:09 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Wed, 12 Dec 2007 15:23:09 -0700
Subject: [SciPy-user] maskedarray
In-Reply-To: <C0C14B2C-2B65-41BB-AEE8-606C938D36C7@enthought.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<200712120924.49482.pgmdevlist@gmail.com>
	<db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
	<C0C14B2C-2B65-41BB-AEE8-606C938D36C7@enthought.com>
Message-ID: <db6b5ecc0712121423s5057838fgaa81924371e239cb@mail.gmail.com>

On Dec 12, 2007 2:21 PM, Travis Vaught <travis at enthought.com> wrote:

> Just to remind folks, there is a scipy channel at irc.freenode.net as
> well...that might be nice to allow more folks to participate, and to
> have an archive of the conversation.

I wasn't even aware of this, as it's not mentioned anywhere I can find
on the site...

I've just added a brief note about it here:

http://scipy.org/Mailing_Lists

Though perhaps someone who uses IRC regularly (I don't) could expand
it with a bit more useful info...

Cheers,

f


From pgmdevlist at gmail.com  Wed Dec 12 17:25:50 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Wed, 12 Dec 2007 17:25:50 -0500
Subject: [SciPy-user] maskedarray
In-Reply-To: <C0C14B2C-2B65-41BB-AEE8-606C938D36C7@enthought.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<db6b5ecc0712121117u2cfec685ieb7e76cd4fbb4da0@mail.gmail.com>
	<C0C14B2C-2B65-41BB-AEE8-606C938D36C7@enthought.com>
Message-ID: <200712121725.52774.pgmdevlist@gmail.com>

On Wednesday 12 December 2007 16:21:29 Travis Vaught wrote:
> On Dec 12, 2007, at 1:17 PM, Fernando Perez wrote:

> Just to remind folks, there is a scipy channel at irc.freenode.net as
> well...that might be nice to allow more folks to participate, and to
> have an archive of the conversation.

Sounds like a good solution indeed, as I don't have Skype. I'll try to stick 
around my box this week-end, let me know what the best times would be.

About the use of maskedarray: so far, no major complaints. It works great with 
matplotlib, it's the base for TimeSeries, it's overall far easier to 
subclass. Of course, parts could be ported to C for optimization, but that 
would bring another set of predicaments. 


From oliphant at enthought.com  Wed Dec 12 18:55:00 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Wed, 12 Dec 2007 17:55:00 -0600
Subject: [SciPy-user] maskedarray
In-Reply-To: <200712120924.49482.pgmdevlist@gmail.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com>
	<200712120924.49482.pgmdevlist@gmail.com>
Message-ID: <476074D4.4080003@enthought.com>

Pierre GM wrote:
> On Tuesday 11 December 2007 15:48:01 Tom Johnson wrote:
>   
>> What is the status regarding 'maskedarray'?  When will this become
>> part (replace ma) of the standard distribution?  Also, what is the
>> recommended way to import sandbox modules so that code changes are
>> minimized when (if) the package becomes part of the standard
>> distribution.
>>     
>
>
> Mmh, no answer from our leaders ? Well, for now maskedarray lives in the scipy 
> sandbox, but it probably could be transferred to the numpy sandbox. There's 
> still no particular decision I'm aware of about putting it the official 
> release. More feedback is probably needed from users.
>   

I apologize for not chiming in here.  It has been difficult to keep up 
with all the lists on which I should ostensibly know what is being 
discussed as I have changed jobs and am settling in to new responsibilites.

It seems to me that it is time to move the scipy.sandbox.maskedarray 
implementation over to numpy.ma.

I'm a little concerned about doing this before 1.1, but if the API has 
not changed, then it should be fine.   If there is no opposition, then 
it seems that this can move forward before 1.0.5 which I would like to 
see in January.

For the future 1.1 release (late 2008), I would like to see (in no 
particular priority order):

1) The MA made a C-subclass
2) The matrix object made a C-subclass (for speed).
3) The sparse matrix structure (but not all the algorithms) brought over 
into NumPy with the indexing matching the matrix indexing.
4) Weave brought over into NumPy

The additions should be created in such a way that they can be disabled 
should so that perhaps a numpy-lite could be produced as well.   Or, 
perhaps better separate packages like numpy-devel, numpy-compat, 
numpy-sparse could be created. 

I'm looking forward to some good discussions over the weekend.

-Travis O.




From s.mientki at ru.nl  Wed Dec 12 19:16:23 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Thu, 13 Dec 2007 01:16:23 +0100
Subject: [SciPy-user] problems with copying / comparing lists,
	containing arrays ?
Message-ID: <476079D7.8050701@ru.nl>

hello,

I've a list, that contains an (numpy) array on one or more locations.
This list acts as a structure to hold all kinds of objects (an array is 
also such an object).
This list is "self.Input_Value" in the code below.
Initially this list is just [ None, None, None, None, None ]
Now some part of the program may fill elements in this list, so it 
become something like this (in practice the array will be much larger)
  [None, array([ 0.        ,  1.77857046,  2.86459363,  2.83518923,  
1.70180685,
  -0.09423228, -1.85357884, -2.89117103, -2.80298683, -1.62336376]), 
None, None, None, None]

I want to test if the array has changed, so I compare it to 
"self.Input_Value_Old",
      if self.Input_Value != self.Input_Value_Old:

If changes are found, I copy Input_Value to Input_Value_Old
      self.Input_Value_Old = copy.copy ( self.Input_Value )

But the next comparison gives an exception,
So the next piece of code

    # Check for any changes
    print 'Check Inputs',self.Input_Value
    print 'Check Inputs',self.Input_Value_Old
    try:
      if self.Input_Value != self.Input_Value_Old:
        print 'INPUTS CHANGED',self.Name
        Change = True
      else :
        print 'INPUT;',self.Name,self.Input_Value
    except:
      print 'ERROR'
    print 'DONE'

Generates the following output:

Check Inputs [None, array([ 0.        ,  1.77857046,  2.86459363,  
2.83518923,  1.70180685,
       -0.09423228, -1.85357884, -2.89117103, -2.80298683, 
-1.62336376]), None, None, None, None]
Check Inputs [None, array([  0.00000000e+00,   1.76335576e+00,   
2.85316955e+00,
         2.85316955e+00,   1.76335576e+00,   3.67381906e-16,
        -1.76335576e+00,  -2.85316955e+00,  -2.85316955e+00,
        -1.76335576e+00]), None, None, None, None]
ERROR
DONE

Each time some value in the array changes, the whole array is generated 
as a new array.
I discovered that if I change the array in place (which is of course the 
final solution),
the exception doesn't occur, but the change is also never detected,
because probably  the next statement is never True
      if self.Input_Value != self.Input_Value_Old:

Now with in place changing of the array, and a deepcopy,
the exception error occurs again and besides that a deepcopy has some 
trouble with other objects.

Is there another way to detect the changes in the array,
or is it better to set a flag when the array is changed ?

thanks,
Stef Mientki



From timmichelsen at gmx-topmail.de  Wed Dec 12 19:35:40 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Thu, 13 Dec 2007 01:35:40 +0100
Subject: [SciPy-user] time series: Python vs. R
Message-ID: <fjpuos$6d9$1@ger.gmane.org>

Hello,
first of all it's nice to hear that there are a few more interested to 
get on with time series.

Unfortunately, I haven't found a lot of time to investigate into the 
packages. Time's kinda short at the end of the year... I just finished 
to install them on my Ubuntu box with checkinstall.

I stumbled onto the R package time series.

It looks like R can really read structured data with ease.

Where would you see the advantage of using Python when analysing time 
series over R? Where is R (still) stronger?

I think for the long run Python will help be more because it would be 
easier to write filters for the measurement data and include it with all 
the other modules available...

Would been nice to hear about your experiences.

Kind regards,
Timmie



From gael.varoquaux at normalesup.org  Wed Dec 12 19:49:20 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Thu, 13 Dec 2007 01:49:20 +0100
Subject: [SciPy-user] time series: Python vs. R
In-Reply-To: <fjpuos$6d9$1@ger.gmane.org>
References: <fjpuos$6d9$1@ger.gmane.org>
Message-ID: <20071213004920.GG29192@clipper.ens.fr>

On Thu, Dec 13, 2007 at 01:35:40AM +0100, Tim Michelsen wrote:
> Where would you see the advantage of using Python when analysing time 
> series over R? Where is R (still) stronger?

Full integration with other tools. With Python you can get very good 2D
plotting, 3D plotting, web service, network abstraction layer, numerical
methods, ...

I use Python in situation where I need a large variety of different
tools. I can indeed agregate different programs throught pipes, shell
calls, .... I've done it before, I am much happier with Python and being
able to put all this in a consistant language rather than a patchwork.

Obviously your demands depend on what you do. However in the long run
there are a strong usecases for all-Python packages, or
well-interface-to-Python packages to implement the different operations
needed in scientific computing.

Cheers,

Ga?l


From aisaac at american.edu  Wed Dec 12 22:27:48 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Wed, 12 Dec 2007 22:27:48 -0500
Subject: [SciPy-user] maskedarray
In-Reply-To: <476074D4.4080003@enthought.com>
References: <b02c60870712111248i7cd6a33bu3b8c1458244cf214@mail.gmail.com><200712120924.49482.pgmdevlist@gmail.com><476074D4.4080003@enthought.com>
Message-ID: <Mahogany-0.67.0-272-20071212-222748.00@american.edu>

On Wed, 12 Dec 2007, "Travis E. Oliphant" apparently wrote:
> 2) The matrix object made a C-subclass (for speed). 

This will probably be the last chance for such a change,
so I again hope that consideration will be given to *one*
change in the matrix object:
iteration over a matrix should return arrays
(instead of matrices).

So if A is a matrix,
A[1] should be an array,
but A[1,:] should be a matrix.

Obviously this is an argument from design rather
than from functionality.  Current behavior is not "natural".

E.g., it makes it awkward to iterate over all elements in
the "natural" way, which I claim is::

    for row in A:
        for element in row:
            print element

This example is just meant to illustrate in a simple way 
what is "odd" about the current behavior.  (It is not meant 
to be an "argument" nor to suggest that the current absence 
simple ways to do the same thing---e.g., using A.A.)
Whenever I am working with matrices, I notice some aspect of 
this "oddity", and it is annoying when so much else is quite 
aesthetic.

Cheers,
Alan Isaac

PS For those who do not know, here is an example of the 
current behavior.  (The following prints 2 matrices instead 
of 4 elements.)

>>> A = N.mat('1 2;3 4')
>>> for row in A:
...     for element in row:
...         print element
...
[[1 2]]
[[3 4]]
>>>





From sgarcia at olfac.univ-lyon1.fr  Thu Dec 13 04:54:24 2007
From: sgarcia at olfac.univ-lyon1.fr (Samuel GARCIA)
Date: Thu, 13 Dec 2007 10:54:24 +0100
Subject: [SciPy-user] scipy.stats.percentileofscore : something strange
Message-ID: <47610150.4090307@olfac.univ-lyon1.fr>

Hi all,
I found something strange in scipy.stats.percentileofscore

In [1]: from scipy import *

In [2]: a = rand(10000)

In [3]: stats.percentileofscore(a,.2)
Out[3]: 20.0157565073
         This OK.


In [4]: stats.percentileofscore(a,.0002)
Out[4]: 102.898311442
            This is strange !!!!!

In [5]: stats.percentileofscore(a,1.4)
---------------------------------------------------------------------------
<type 'exceptions.IndexError'>            Traceback (most recent call last)

/home/sgarcia/<ipython console> in <module>()

/usr/lib/python2.5/site-packages/scipy/stats/stats.py in
percentileofscore(a, score, histbins, defaultlimits)
    942     cumhist = np.cumsum(h*1, axis=0)
    943     i = int((score - lrl)/float(binsize))
--> 944     pct =
(cumhist[i-1]+((score-(lrl+binsize*i))/float(binsize))*h[i])/float(len(a))
* 100
    945     return pct
    946
<type 'exceptions.IndexError'>: index out of bounds

               This does not work...

Any idea, why ?

Sam

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Samuel Garcia
Laboratoire de Neurosciences Sensorielles, Comportement, Cognition.
CNRS - UMR5020 - Universite Claude Bernard LYON 1
Equipe logistique et technique
50, avenue Tony Garnier
69366 LYON Cedex 07
FRANCE
T?l : 04 37 28 74 64
Fax : 04 37 28 76 01
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071213/1f2d99ef/attachment.html>

From matthieu.brucher at gmail.com  Thu Dec 13 05:24:09 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 13 Dec 2007 11:24:09 +0100
Subject: [SciPy-user] scipy.stats.percentileofscore : something strange
In-Reply-To: <47610150.4090307@olfac.univ-lyon1.fr>
References: <47610150.4090307@olfac.univ-lyon1.fr>
Message-ID: <e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com>

2007/12/13, Samuel GARCIA <sgarcia at olfac.univ-lyon1.fr>:
>
>  Hi all,
> I found something strange in scipy.stats.percentileofscore
> In [1]: from scipy import *
> In [2]: a = rand(10000)
> In [3]: stats.percentileofscore(a,.2)
> Out[3]: 20.0157565073
>          This OK.
> In [4]: stats.percentileofscore(a,.0002)
> Out[4]: 102.898311442
>             This is strange !!!!!
>
>
This may be a bug


In [5]: stats.percentileofscore(a,1.4)
> ---------------------------------------------------------------------------
> <type 'exceptions.IndexError'>            Traceback (most recent call last)
> */home/sgarcia/*<ipython console> in <module>()
> /usr/lib/python2.5/site-packages/scipy/stats/stats.py in
> percentileofscore(a, score, histbins, defaultlimits)
>     942     cumhist = np.cumsum(h*1, axis=0)
>     943     i = int((score - lrl)/float(binsize))
> --> 944     pct =
> (cumhist[i-1]+((score-(lrl+binsize*i))/float(binsize))*h[i])/float(len(a))
> * 100
>     945     return pct
>     946<type 'exceptions.IndexError'>: index out of bounds
>                This does not work...
>
>
I expect this to raise this exception, you cannot take a percentile outside
[0, 1] (1.4 = 140%)

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071213/a5cdad9e/attachment.html>

From sgarcia at olfac.univ-lyon1.fr  Thu Dec 13 05:36:15 2007
From: sgarcia at olfac.univ-lyon1.fr (Samuel GARCIA)
Date: Thu, 13 Dec 2007 11:36:15 +0100
Subject: [SciPy-user] scipy.stats.percentileofscore : something strange
In-Reply-To: <e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com>
References: <47610150.4090307@olfac.univ-lyon1.fr>
	<e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com>
Message-ID: <47610B1F.5010502@olfac.univ-lyon1.fr>



Matthieu Brucher a ?crit :
>
>
> 2007/12/13, Samuel GARCIA <sgarcia at olfac.univ-lyon1.fr 
> <mailto:sgarcia at olfac.univ-lyon1.fr>>:
>
>     Hi all,
>     I found something strange in scipy.stats.percentileofscore
>
>     In [1]: from scipy import *
>
>     In [2]: a = rand(10000)
>
>     In [3]: stats.percentileofscore(a,.2)
>     Out[3]: 20.0157565073
>              This OK.
>
>
>     In [4]: stats.percentileofscore(a,.0002)
>     Out[4]: 102.898311442
>                 This is strange !!!!!
>
>
> This may be a bug
>  
>
>     In [5]: stats.percentileofscore(a,1.4)
>     ---------------------------------------------------------------------------
>     <type 'exceptions.IndexError
>     '>            Traceback (most recent call last)
>
>     //home/sgarcia//<ipython console> in <module>()
>
>     /usr/lib/python2.5/site-packages/scipy/stats/stats.py in
>     percentileofscore(a, score, histbins, defaultlimits)
>         942     cumhist = np.cumsum(h*1, axis=0)
>         943     i = int((score - lrl)/float(binsize))
>
>     --> 944     pct =
>     (cumhist[i-1]+((score-(lrl+binsize*i))/float(binsize))*h[i])/float(len(a))
>     * 100
>         945     return pct
>         946
>
>     <type 'exceptions.IndexError'>: index out of bounds
>
>                    This does not work...
>
>
> I expect this to raise this exception, you cannot take a percentile 
> outside [0, 1] (1.4 = 140%)

I think that it should return 100 if superior to rigtht limit and 0 if 
inferior to left limit.
No ?


>
> Matthieu
> -- 
> French PhD student
> Website : http://matthieu-brucher.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
>   

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Samuel Garcia
Laboratoire de Neurosciences Sensorielles, Comportement, Cognition.
CNRS - UMR5020 - Universite Claude Bernard LYON 1
Equipe logistique et technique
50, avenue Tony Garnier
69366 LYON Cedex 07
FRANCE
T?l : 04 37 28 74 64
Fax : 04 37 28 76 01
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071213/c9f82d79/attachment.html>

From xavier.barthelemy at cmla.ens-cachan.fr  Thu Dec 13 05:48:43 2007
From: xavier.barthelemy at cmla.ens-cachan.fr (Xavier Barthelemy)
Date: Thu, 13 Dec 2007 11:48:43 +0100
Subject: [SciPy-user] problem in reading a indices
In-Reply-To: <e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com>
References: <47610150.4090307@olfac.univ-lyon1.fr>
	<e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com>
Message-ID: <47610E0B.20502@cmla.ens-cachan.fr>

Hi all

I'm becoming mad, because I can't see what's wrong:

I am constructing a GUI, to plot some datas.
so let's have a look of what's wrong:



in my code I have a variable named choice[i].current which is the 
current selection of the i-th Listbox object. it is a tuple, with one 
element.

so when I write

print type(i),type(choice[i].current)
I have: int and tuple

print type(i),type(choice[i].current[0])
I have: int and str

print type(i),type(int(choice[i].current[0]))
I have: int and int

so when I call another array with these indices
ArrayWithData[i,int(choice[i].current[0])]

I have the following error: TypeError: list indices must be integers

so I tried an intermediate value, because sometimes, the oneliner code 
doesn't work, so with an intermediate passage: 
value=int(choice[i].current[0])
ArrayWithData[i,value]

I have the same error

and I don't understand why. What's wrong?
May anyone have an idea?
Xavier

pm:
and print type(ArrayWithData), ArrayWithData gives me
<type 'list'> [array([ 2.01,  5.01]),...]




From rgold at lsw.uni-heidelberg.de  Thu Dec 13 05:20:57 2007
From: rgold at lsw.uni-heidelberg.de (rgold at lsw.uni-heidelberg.de)
Date: Thu, 13 Dec 2007 11:20:57 +0100 (CET)
Subject: [SciPy-user] read ascii data
In-Reply-To: <200712121157.14457.cscheit@lstm.uni-erlangen.de>
References: <mailman.10357.1197403323.2301.scipy-user@scipy.org>
	<200712121157.14457.cscheit@lstm.uni-erlangen.de>
Message-ID: <36424.147.142.111.40.1197541257.squirrel@srv0.lsw.uni-heidelberg.de>

Hi Christoph,

Probably you restrict yourself to read only certain rows of your ascii
file because of performance issues... If this is not the case you might
want to ignore this Email!

Note, that if these lines are at the very end of your ascii-file you don't
gain much speed, because there's no line end in an ascii-file (most of the
file has to be read anyway). So if you know the size of your array,
depending on what you want to do it's maybe faster to read in the whole
ascii-file as one single string and then selecting the lines you want
using split(). i.e.:

LIST=[float(D) for D in FILE.read().split()]

It is simple, but very fast! However I suggest to store your data in
binary format, it's actually very simple handling them using python.

Cheers,
Roman

> Hi All,
>
> I have a file containing ascii data related to a block-structured grid.
> I could use readtxt, but I would like to have the possibility to read only
> a certain range of rows out of the file. Lets say I have ten blocks in my
> grid
> and each block 10 rows of data.
> So in one read I would like to read for instance line 30 to line 39
> I can use skip rows, but how can I limit the loadtxt to the upper limit?
> Is there something already available or do I have to do it on my own?
>
> Thanks in advance,
>
> Christoph
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>




From matthieu.brucher at gmail.com  Thu Dec 13 05:53:40 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 13 Dec 2007 11:53:40 +0100
Subject: [SciPy-user] scipy.stats.percentileofscore : something strange
In-Reply-To: <47610B1F.5010502@olfac.univ-lyon1.fr>
References: <47610150.4090307@olfac.univ-lyon1.fr>
	<e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com>
	<47610B1F.5010502@olfac.univ-lyon1.fr>
Message-ID: <e76aa17f0712130253x605ed760g7e37fb1e0030e4cb@mail.gmail.com>

2007/12/13, Samuel GARCIA <sgarcia at olfac.univ-lyon1.fr>:
>
>
>
> Matthieu Brucher a ?crit :
>
>
>
> 2007/12/13, Samuel GARCIA <sgarcia at olfac.univ-lyon1.fr>:
> >
> >  Hi all,
> > I found something strange in scipy.stats.percentileofscore
> > In [1]: from scipy import *
> > In [2]: a = rand(10000)
> > In [3]: stats.percentileofscore(a,.2)
> > Out[3]: 20.0157565073
> >          This OK.
> > In [4]: stats.percentileofscore(a,.0002)
> > Out[4]: 102.898311442
> >             This is strange !!!!!
> >
> >
> This may be a bug
>
>
If you put 0 for the score, a number greater than 100 is returned, so this
is a bug. But for 0.0002, your percentile will be null, you can change the
histbins value (default=10) for a better precision.


I expect this to raise this exception, you cannot take a percentile outside
> [0, 1] (1.4 = 140%)
>
>
> I think that it should return 100 if superior to rigtht limit and 0 if
> inferior to left limit.
> No ?
>


My mistake, I was confused by the interval. I think you are right, but
perhaps someone else will give the answer to this behaviour.
I think you should open a ticket on the TRAC to keep track of this bug.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071213/7d20687e/attachment.html>

From nwagner at iam.uni-stuttgart.de  Thu Dec 13 07:06:47 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 13 Dec 2007 13:06:47 +0100
Subject: [SciPy-user] Rutherford Boeing format
Message-ID: <web-93275518@uni-stuttgart.de>

Hi all,

I was wondering if someone has written a function to 
import matrices given in the Rutherford Boeing format.

Nils

Reference:
http://www.cise.ufl.edu/research/sparse/RBio/


From nwagner at iam.uni-stuttgart.de  Thu Dec 13 07:57:25 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 13 Dec 2007 13:57:25 +0100
Subject: [SciPy-user] Creating a sparse matrix
Message-ID: <web-93277839@uni-stuttgart.de>

Hi all,

How can I build a sparse matrix from the following array ?
>>> data
array([['1', '1', '1.7244067583090530E+05'],
        ['1', '2', '4.7526228631699840E+04'],
        ['2', '2', '1.7244064192412970E+05'],
        ['3', '3', '2.7456944493758570E+02'],
        ['1', '4', '2.3763114315849920E+05'],
        ['2', '4', '8.6220320962064820E+05'],
        ['3', '4', '5.9378661309939430E+02'],
        ['4', '4', '4.3948451437807490E+06'],
        ['1', '5', '-8.6220337915452610E+05'],
        ['2', '5', '-2.3763114315849920E+05'],
        ['3', '5', '-5.9378754413417760E+02'],
        ['4', '5', '-1.1952119204729520E+06'],
        ['5', '5', '4.3948459988011970E+06'],
        ['6', '6', '4.2234547275787270E+03'],
        ['1', '7', '-1.6598232969237920E+04'],
        ['2', '7', '-6.4128987931960360E+03'],
        ['4', '7', '-3.2064493965980180E+04'],
        ['5', '7', '8.2991164846189540E+04'],
        ['7', '7', '1.7244065067439350E+05'],
        ['1', '8', '6.4129110657815880E+03'],
        ['2', '8', '-6.0501784478906850E+04'],
        ['4', '8', '-3.0250892239453440E+05'],
        ['5', '8', '-3.2064555328907960E+04'],
        ['7', '8', '-4.7526234795758550E+04'],
        ['8', '8', '1.7244065036281190E+05'],
        ['3', '9', '-6.6701524177624400E+01'],
        ['4', '9', '1.0014969892580470E+03'],
        ['5', '9', '1.8006472158568300E+02'],
        ['9', '9', '2.7456950540497500E+02'],
        ['1', '10', '3.2064555328907960E+04'],
        ['2', '10', '-3.0250892239453440E+05'],
        ['3', '10', '-1.0014968431605310E+03'],
        ['4', '10', '-1.4981909287503440E+06'],
        ['5', '10', '-1.5690996463630880E+05'],
        ['7', '10', '-2.3763117397879280E+05'],
        ['8', '10', '8.6220325181405960E+05'],
        ['9', '10', '-5.9378786697859830E+02'],
        ['10', '10', '4.3948453583907450E+06'],
        ['1', '11', '8.2991164846189560E+04'],
        ['2', '11', '3.2064493965980140E+04'],
        ['3', '11', '1.8006434238808650E+02'],
        ['4', '11', '1.5690966299256120E+05'],
        ['5', '11', '-4.2216699312324380E+05'],
        ['7', '11', '-8.6220325337196720E+05'],
        ['8', '11', '2.3763117397879260E+05'],
        ['9', '11', '-5.9378773737342270E+02'],
        ['10', '11', '1.1952120829330800E+06'],
        ['11', '11', '4.3948453666102750E+06'],
        ['6', '12', '-1.0245631116652000E+03'],
        ['12', '12', '4.2234546752310640E+03'],
        ['1', '13', '-2.1562536183412980E+04'],
        ['2', '13', '-1.3326854263605380E+04'],
        ['4', '13', '-6.6634271318026890E+04'],
        ['5', '13', '1.0781268091706490E+05'],
        ['7', '13', '-6.0501787019015210E+04'],
        ['8', '13', '-6.4129083147729780E+03'],
        ['10', '13', '-3.2064541573864920E+04'],
        ['11', '13', '3.0250893509507610E+05'],
        ['13', '13', '1.7244065108409200E+05'],
        ['1', '14', '-1.3326856939548480E+04'],
        ['2', '14', '-2.1562538882437520E+04'],
        ['4', '14', '-1.0781269441218760E+05'],
        ['5', '14', '6.6634284697742330E+04'],
        ['7', '14', '6.4129066592753080E+03'],
        ['8', '14', '-1.6598230710213180E+04'],
        ['10', '14', '-8.2991153551065850E+04'],
        ['11', '14', '-3.2064533296376550E+04'],
        ['13', '14', '4.7526232819775960E+04'],
        ['14', '14', '1.7244065558155810E+05'],
        ['3', '15', '3.9071327498923200E+00'],
        ['4', '15', '1.1208665237160810E+02'],
        ['5', '15', '-1.1208648718752250E+02'],
        ['9', '15', '-6.6701524986800910E+01'],
        ['10', '15', '1.8006454779629130E+02'],
        ['11', '15', '-1.0014969364802770E+03'],
        ['15', '15', '2.7456952836064130E+02'],
        ['1', '16', '-6.6634284697742330E+04'],
        ['2', '16', '-1.0781269441218760E+05'],
        ['3', '16', '-1.1208653552827830E+02'],
        ['4', '16', '-5.3738372531588020E+05'],
        ['5', '16', '3.3136527542132390E+05'],
        ['7', '16', '3.2064533296376570E+04'],
        ['8', '16', '-8.2991153551065860E+04'],
        ['9', '16', '1.8006450558919260E+02'],
        ['10', '16', '-4.2216693696279110E+05'],
        ['11', '16', '-1.5690985700615200E+05'],
        ['13', '16', '2.3763116409887990E+05'],
        ['14', '16', '8.6220327790779050E+05'],
        ['15', '16', '-5.9378754489808300E+02'],
        ['16', '16', '4.3948454932894830E+06'],
        ['1', '17', '1.0781268091706490E+05'],
        ['2', '17', '6.6634271318026870E+04'],
        ['3', '17', '1.1208660539636940E+02'],
        ['4', '17', '3.3136520639089480E+05'],
        ['5', '17', '-5.3738365971501340E+05'],
        ['7', '17', '3.0250893509507610E+05'],
        ['8', '17', '3.2064541573864890E+04'],
        ['9', '17', '1.0014969080580460E+03'],
        ['10', '17', '1.5690989812553410E+05'],
        ['11', '17', '-1.4981909926673520E+06'],
        ['13', '17', '-8.6220325542046030E+05'],
        ['14', '17', '-2.3763116409887980E+05'],
        ['15', '17', '5.9378787780320290E+02'],
        ['16', '17', '-1.1952120315502540E+06'],
        ['17', '17', '4.3948453781456250E+06'],
        ['6', '18', '-4.0834357736871400E+02'],
        ['12', '18', '-1.0245631207891270E+03'],
        ['18', '18', '4.2234546579613690E+03'],
        ['1', '19', '-6.0501796429481100E+04'],
        ['2', '19', '6.4128982934224380E+03'],
        ['4', '19', '3.2064491467112190E+04'],
        ['5', '19', '3.0250898214740550E+05'],
        ['7', '19', '-2.1562538882438510E+04'],
        ['8', '19', '1.3326853882609100E+04'],
        ['10', '19', '6.6634269413045520E+04'],
        ['11', '19', '1.0781269441219260E+05'],
        ['13', '19', '-1.6598239767392180E+04'],
        ['14', '19', '-6.4129061595026810E+03'],
        ['16', '19', '-3.2064530797513400E+04'],
        ['17', '19', '8.2991198836960840E+04'],
        ['19', '19', '1.7244064683129310E+05'],
        ['1', '20', '-6.4129127399249150E+03'],
        ['2', '20', '-1.6598242026418230E+04'],
        ['4', '20', '-8.2991210132091170E+04'],
        ['5', '20', '3.2064563699624550E+04'],
        ['7', '20', '1.3326857320545740E+04'],
        ['8', '20', '-2.1562536183413320E+04'],
        ['10', '20', '-1.0781268091706660E+05'],
        ['11', '20', '-6.6634286602728680E+04'],
        ['13', '20', '6.4129099889152090E+03'],
        ['14', '20', '-6.0501798969589890E+04'],
        ['16', '20', '-3.0250899484794950E+05'],
        ['17', '20', '-3.2064549944576080E+04'],
        ['19', '20', '-4.7526226655715920E+04'],
        ['20', '20', '1.7244067655218550E+05'],
        ['3', '21', '-6.6701552188476970E+01'],
        ['4', '21', '-1.8006429542064950E+02'],
        ['5', '21', '-1.0014970368071590E+03'],
        ['9', '21', '3.9071327498907450E+00'],
        ['10', '21', '-1.1208655989768520E+02'],
        ['11', '21', '-1.1208657177767270E+02'],
        ['15', '21', '-6.6701552997659000E+01'],
        ['16', '21', '-1.0014972526987650E+03'],
        ['17', '21', '-1.8006453997854880E+02'],
        ['21', '21', '2.7456946789324640E+02'],
        ['1', '22', '-3.2064563699624570E+04'],
        ['2', '22', '-8.2991210132091170E+04'],
        ['3', '22', '-1.8006471376798420E+02'],
        ['4', '22', '-4.2216722298368720E+05'],
        ['5', '22', '1.5691000662161880E+05'],
        ['7', '22', '6.6634286602728650E+04'],
        ['8', '22', '-1.0781268091706660E+05'],
        ['9', '22', '1.1208653268623420E+02'],
        ['10', '22', '-5.3738365971502150E+05'],
        ['11', '22', '-3.3136528414261740E+05'],
        ['13', '22', '3.2064549944576050E+04'],
        ['14', '22', '-3.0250899484794950E+05'],
        ['15', '22', '1.0014971017047370E+03'],
        ['16', '22', '-1.4981912910366350E+06'],
        ['17', '22', '-1.5690994011081730E+05'],
        ['19', '22', '-2.3763113327857960E+05'],
        ['20', '22', '8.6220338276092740E+05'],
        ['21', '22', '5.9378755495883890E+02'],
        ['22', '22', '4.3948460185560810E+06'],
        ['1', '23', '3.0250898214740550E+05'],
        ['2', '23', '-3.2064491467112220E+04'],
        ['3', '23', '1.0014973054764940E+03'],
        ['4', '23', '-1.5690965052770850E+05'],
        ['5', '23', '-1.4981912271196160E+06'],
        ['7', '23', '1.0781269441219260E+05'],
        ['8', '23', '-6.6634269413045550E+04'],
        ['9', '23', '1.1208661612225140E+02'],
        ['10', '23', '-3.3136519766962600E+05'],
        ['11', '23', '-5.3738372531590490E+05'],
        ['13', '23', '8.2991198836960880E+04'],
        ['14', '23', '3.2064530797513400E+04'],
        ['15', '23', '-1.8006445862190810E+02'],
        ['16', '23', '1.5690984454132130E+05'],
        ['17', '23', '-4.2216716682320250E+05'],
        ['19', '23', '-8.6220323415646550E+05'],
        ['20', '23', '2.3763113327857960E+05'],
        ['21', '23', '5.9378642062413570E+02'],
        ['22', '23', '1.1952118690900900E+06'],
        ['23', '23', '4.3948452704599260E+06'],
        ['6', '24', '-1.0245630840369950E+03'],
        ['12', '24', '-4.0834357736871400E+02'],
        ['18', '24', '-1.0245630931609220E+03'],
        ['24', '24', '4.2234547103090340E+03']],
       dtype='|S23')

data contains information about row, column and the 
corresponding entry.

Any pointer would be appreciated.

Thanks in advance
                       Nils


From cimrman3 at ntc.zcu.cz  Thu Dec 13 08:03:14 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Thu, 13 Dec 2007 14:03:14 +0100
Subject: [SciPy-user] Creating a sparse matrix
In-Reply-To: <web-93277839@uni-stuttgart.de>
References: <web-93277839@uni-stuttgart.de>
Message-ID: <47612D92.4020408@ntc.zcu.cz>

Nils Wagner wrote:
> Hi all,
> 
> How can I build a sparse matrix from the following array ?
>>>> data
> array([['1', '1', '1.7244067583090530E+05'],
>         ['1', '2', '4.7526228631699840E+04'],
 > ...
>         ['18', '24', '-1.0245630931609220E+03'],
>         ['24', '24', '4.2234547103090340E+03']],
>        dtype='|S23')
> 
> data contains information about row, column and the 
> corresponding entry.
> 
> Any pointer would be appreciated.

You read the data from a file, right? You should convert the row/column 
indices to integers and values to floats during the reading. Then you 
would have your data ready for the COO matrix constructor.

r.



From nwagner at iam.uni-stuttgart.de  Thu Dec 13 08:12:51 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 13 Dec 2007 14:12:51 +0100
Subject: [SciPy-user] Creating a sparse matrix
In-Reply-To: <47612D92.4020408@ntc.zcu.cz>
References: <web-93277839@uni-stuttgart.de>
 <47612D92.4020408@ntc.zcu.cz>
Message-ID: <web-93278688@uni-stuttgart.de>

On Thu, 13 Dec 2007 14:03:14 +0100
  Robert Cimrman <cimrman3 at ntc.zcu.cz> wrote:
> Nils Wagner wrote:
>> Hi all,
>> 
>> How can I build a sparse matrix from the following array 
>>?
>>>>> data
>> array([['1', '1', '1.7244067583090530E+05'],
>>         ['1', '2', '4.7526228631699840E+04'],
> > ...
>>         ['18', '24', '-1.0245630931609220E+03'],
>>         ['24', '24', '4.2234547103090340E+03']],
>>        dtype='|S23')
>> 
>> data contains information about row, column and the 
>> corresponding entry.
>> 
>> Any pointer would be appreciated.
> 
> You read the data from a file, right? 

Exactly.

You should convert
>the row/column 
> indices to integers and values to floats during the 
>reading. 

How can I do that on-the-fly ?

Then you
> would have your data ready for the COO matrix 
>constructor.

Nils
  


From aisaac at american.edu  Thu Dec 13 08:13:55 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Thu, 13 Dec 2007 08:13:55 -0500
Subject: [SciPy-user] problem in reading a indices
In-Reply-To: <47610E0B.20502@cmla.ens-cachan.fr>
References: <47610150.4090307@olfac.univ-lyon1.fr><e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com><47610E0B.20502@cmla.ens-cachan.fr>
Message-ID: <Mahogany-0.67.0-272-20071213-081355.00@american.edu>

On Thu, 13 Dec 2007, Xavier Barthelemy apparently wrote:
> ArrayWithData[i,int(choice[i].current[0])] I have the 
> following error: TypeError: list indices must be integers 

Looks like ArrayWithData is a list, not an array.

Cheers,
Alan Isaac





From aisaac at american.edu  Thu Dec 13 08:16:41 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Thu, 13 Dec 2007 08:16:41 -0500
Subject: [SciPy-user] problem in reading a indices
In-Reply-To: <47610E0B.20502@cmla.ens-cachan.fr>
References: <47610150.4090307@olfac.univ-lyon1.fr><e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com><47610E0B.20502@cmla.ens-cachan.fr>
Message-ID: <Mahogany-0.67.0-272-20071213-081641.00@american.edu>

On Thu, 13 Dec 2007, Xavier Barthelemy apparently wrote:
> ArrayWithData[i,int(choice[i].current[0])] 
> pm:
> and print type(ArrayWithData), ArrayWithData gives me 
> <type 'list'> [array([ 2.01,  5.01]),...] 

Sorry, missed your postscript.  (What does pm mean?)
So ArrayWithData is a list of arrays.  Then try
    ArrayWithData[i][otherindex]
since lists do not handle multiple indexes.

Cheers,
Alan Isaac





From aisaac at american.edu  Thu Dec 13 08:26:06 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Thu, 13 Dec 2007 08:26:06 -0500
Subject: [SciPy-user] Creating a sparse matrix
In-Reply-To: <web-93278688@uni-stuttgart.de>
References: <web-93277839@uni-stuttgart.de><47612D92.4020408@ntc.zcu.cz><web-93278688@uni-stuttgart.de>
Message-ID: <Mahogany-0.67.0-928-20071213-082606.00@american.edu>

On Thu, 13 Dec 2007, Nils Wagner apparently wrote:
> How can I do that on-the-fly ? 

That depends on the file format.
But are you just asking if you can do the following?

data = []
for row in open(filename,'r'):
    r0, r1, r2 = row.split()
    data.append( [int(r0), int(r1), float(r2)] )

Cheers,
Alan Isaac





From cimrman3 at ntc.zcu.cz  Thu Dec 13 08:27:05 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Thu, 13 Dec 2007 14:27:05 +0100
Subject: [SciPy-user] Creating a sparse matrix
In-Reply-To: <web-93278688@uni-stuttgart.de>
References: <web-93277839@uni-stuttgart.de> <47612D92.4020408@ntc.zcu.cz>
	<web-93278688@uni-stuttgart.de>
Message-ID: <47613329.4020603@ntc.zcu.cz>

Nils Wagner wrote:
> On Thu, 13 Dec 2007 14:03:14 +0100
>   Robert Cimrman <cimrman3 at ntc.zcu.cz> wrote:
>> Nils Wagner wrote:
>>> Hi all,
>>>
>>> How can I build a sparse matrix from the following array 
>>> ?
>>>>>> data
>>> array([['1', '1', '1.7244067583090530E+05'],
>>>         ['1', '2', '4.7526228631699840E+04'],
>>> ...
>>>         ['18', '24', '-1.0245630931609220E+03'],
>>>         ['24', '24', '4.2234547103090340E+03']],
>>>        dtype='|S23')
>>>
>>> data contains information about row, column and the 
>>> corresponding entry.
>>>
>>> Any pointer would be appreciated.
>> You read the data from a file, right? 
> 
> Exactly.
> 
> You should convert
>> the row/column 
>> indices to integers and values to floats during the 
>> reading. 
> 
> How can I do that on-the-fly ?

Well, supposing the file is not too large:
     fd = open( ... )
     rows, cols, vals = [], [], []
     while 1:
         try:
             line = fd.readline()
             if (len( line ) == 0): break
             if len( line ) == 1: continue
         except EOFError:
             break
         line = line.split()
         ir, ic, val = int( line[0] ), int( line[1] ), float( line[2] )
         rows.append( ir )
         cols.append( ic )
         vals.append( val )

Alternatively, you can try to use numpy.fromfile( ..., sep = ' ', dtype 
= numpy.float64 ) and then reshape it to (n, 3) and re-type the first 
two columns to ints.

r.


From xavier.barthelemy at cmla.ens-cachan.fr  Thu Dec 13 09:10:51 2007
From: xavier.barthelemy at cmla.ens-cachan.fr (Xavier Barthelemy)
Date: Thu, 13 Dec 2007 15:10:51 +0100
Subject: [SciPy-user] problem in reading a indices
In-Reply-To: <Mahogany-0.67.0-272-20071213-081641.00@american.edu>
References: <47610150.4090307@olfac.univ-lyon1.fr><e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com><47610E0B.20502@cmla.ens-cachan.fr>
	<Mahogany-0.67.0-272-20071213-081641.00@american.edu>
Message-ID: <47613D6B.8060509@cmla.ens-cachan.fr>

Yesss   it's working!!!

thanks, i did pass too much time on it for this

thanks a lot. I did not know this point in calling lists.

and by the way, pm is a latinicism for: post mailum ;)

Xavier

Alan G Isaac a ?crit :
> On Thu, 13 Dec 2007, Xavier Barthelemy apparently wrote:
>> ArrayWithData[i,int(choice[i].current[0])] 
>> pm:
>> and print type(ArrayWithData), ArrayWithData gives me 
>> <type 'list'> [array([ 2.01,  5.01]),...] 
> 
> Sorry, missed your postscript.  (What does pm mean?)
> So ArrayWithData is a list of arrays.  Then try
>     ArrayWithData[i][otherindex]
> since lists do not handle multiple indexes.
> 



From nwagner at iam.uni-stuttgart.de  Thu Dec 13 09:27:41 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 13 Dec 2007 15:27:41 +0100
Subject: [SciPy-user] Creating a sparse matrix
In-Reply-To: <47613329.4020603@ntc.zcu.cz>
References: <web-93277839@uni-stuttgart.de> <47612D92.4020408@ntc.zcu.cz>
	<web-93278688@uni-stuttgart.de> <47613329.4020603@ntc.zcu.cz>
Message-ID: <web-93282502@uni-stuttgart.de>

On Thu, 13 Dec 2007 14:27:05 +0100
  Robert Cimrman <cimrman3 at ntc.zcu.cz> wrote:
> Nils Wagner wrote:
>> On Thu, 13 Dec 2007 14:03:14 +0100
>>   Robert Cimrman <cimrman3 at ntc.zcu.cz> wrote:
>>> Nils Wagner wrote:
>>>> Hi all,
>>>>
>>>> How can I build a sparse matrix from the following array 
>>>> ?
>>>>>>> data
>>>> array([['1', '1', '1.7244067583090530E+05'],
>>>>         ['1', '2', '4.7526228631699840E+04'],
>>>> ...
>>>>         ['18', '24', '-1.0245630931609220E+03'],
>>>>         ['24', '24', '4.2234547103090340E+03']],
>>>>        dtype='|S23')
>>>>
>>>> data contains information about row, column and the 
>>>> corresponding entry.
>>>>
>>>> Any pointer would be appreciated.
>>> You read the data from a file, right? 
>> 
>> Exactly.
>> 
>> You should convert
>>> the row/column 
>>> indices to integers and values to floats during the 
>>> reading. 
>> 
>> How can I do that on-the-fly ?
> 
> Well, supposing the file is not too large:
>     fd = open( ... )
>     rows, cols, vals = [], [], []
>     while 1:
>         try:
>             line = fd.readline()
>             if (len( line ) == 0): break
>             if len( line ) == 1: continue
>         except EOFError:
>             break
>         line = line.split()
>         ir, ic, val = int( line[0] ), int( line[1] ), 
>float( line[2] )
>         rows.append( ir )
>         cols.append( ic )
>         vals.append( val )
> 
> Alternatively, you can try to use numpy.fromfile( ..., 
>sep = ' ', dtype 
> = numpy.float64 ) and then reshape it to (n, 3) and 
>re-type the first 
> two columns to ints.
> 
> r.
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
  
Hi Robert,

Thank you very much for your code snippet !
How do I use the COO constructor to build the
matrix from rows, cols and vals ?

Cheers,
             Nils


From strang at nmr.mgh.harvard.edu  Thu Dec 13 09:56:35 2007
From: strang at nmr.mgh.harvard.edu (Gary Strangman)
Date: Thu, 13 Dec 2007 09:56:35 -0500 (EST)
Subject: [SciPy-user] scipy.stats.percentileofscore : something strange
In-Reply-To: <47610B1F.5010502@olfac.univ-lyon1.fr>
References: <47610150.4090307@olfac.univ-lyon1.fr>
	<e76aa17f0712130224j49c18311sff834feeef5b2d94@mail.gmail.com>
	<47610B1F.5010502@olfac.univ-lyon1.fr>
Message-ID: <Pine.LNX.4.62.0712130951180.11007@gate.nmr.mgh.harvard.edu>


Yes, this is a bug. I believe there is no checking for edge or out of 
bound cases. (At least there wasn't when it was first added to scipy ...)

Gary

On Thu, 13 Dec 2007, Samuel GARCIA wrote:

>
>
> Matthieu Brucher a ?crit :
>> 
>> 
>> 2007/12/13, Samuel GARCIA <sgarcia at olfac.univ-lyon1.fr 
>> <mailto:sgarcia at olfac.univ-lyon1.fr>>:
>> 
>>     Hi all,
>>     I found something strange in scipy.stats.percentileofscore
>> 
>>     In [1]: from scipy import *
>> 
>>     In [2]: a = rand(10000)
>> 
>>     In [3]: stats.percentileofscore(a,.2)
>>     Out[3]: 20.0157565073
>>              This OK.
>> 
>> 
>>     In [4]: stats.percentileofscore(a,.0002)
>>     Out[4]: 102.898311442
>>                 This is strange !!!!!
>> 
>> 
>> This may be a bug
>> 
>>     In [5]: stats.percentileofscore(a,1.4)
>>     ---------------------------------------------------------------------------
>>     <type 'exceptions.IndexError
>>     '>            Traceback (most recent call last)
>> 
>>     //home/sgarcia//<ipython console> in <module>()
>> 
>>     /usr/lib/python2.5/site-packages/scipy/stats/stats.py in
>>     percentileofscore(a, score, histbins, defaultlimits)
>>         942     cumhist = np.cumsum(h*1, axis=0)
>>         943     i = int((score - lrl)/float(binsize))
>> 
>>     --> 944     pct =
>>     (cumhist[i-1]+((score-(lrl+binsize*i))/float(binsize))*h[i])/float(len(a))
>>     * 100
>>         945     return pct
>>         946
>> 
>>     <type 'exceptions.IndexError'>: index out of bounds
>> 
>>                    This does not work...
>> 
>> 
>> I expect this to raise this exception, you cannot take a percentile outside 
>> [0, 1] (1.4 = 140%)
>
> I think that it should return 100 if superior to rigtht limit and 0 if 
> inferior to left limit.
> No ?
>
>
>> 
>> Matthieu
>> -- 
>> French PhD student
>> Website : http://matthieu-brucher.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
>> 
>
>

From dominique.orban at gmail.com  Thu Dec 13 10:03:58 2007
From: dominique.orban at gmail.com (Dominique Orban)
Date: Thu, 13 Dec 2007 10:03:58 -0500
Subject: [SciPy-user] Rutherford Boeing format
In-Reply-To: <web-93275518@uni-stuttgart.de>
References: <web-93275518@uni-stuttgart.de>
Message-ID: <8793ae6e0712130703k4e6f2a87qcab343a67841748b@mail.gmail.com>

On 12/13/07, Nils Wagner <nwagner at iam.uni-stuttgart.de> wrote:
> Hi all,
>
> I was wondering if someone has written a function to
> import matrices given in the Rutherford Boeing format.

A while ago I wrote a function to import Harwell-Boeing format. RB is
quite similar so I expect few changed will do the trick. My function
relies on Konrad Hinsen's FortranFormat module, which I also attach
below. I haven't looked at this lately, so I hope it works for you. If
this is useful, I have no problem including it in SciPy.

In the test function, the sparsetools module is part of NLPy and plots
a sparsity pattern. If you don't use NLPy, just replace that with any
other plot you like :)

Dominique
-------------- next part --------------
A non-text attachment was scrubbed...
Name: FortranFormat.py
Type: text/x-python-script
Size: 6117 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071213/788a9027/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: hb.py
Type: text/x-python-script
Size: 2848 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071213/788a9027/attachment-0001.bin>

From cimrman3 at ntc.zcu.cz  Thu Dec 13 10:45:43 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Thu, 13 Dec 2007 16:45:43 +0100
Subject: [SciPy-user] Creating a sparse matrix
In-Reply-To: <web-93282502@uni-stuttgart.de>
References: <web-93277839@uni-stuttgart.de>
	<47612D92.4020408@ntc.zcu.cz>	<web-93278688@uni-stuttgart.de>
	<47613329.4020603@ntc.zcu.cz> <web-93282502@uni-stuttgart.de>
Message-ID: <476153A7.1020707@ntc.zcu.cz>

Nils Wagner wrote:
> Thank you very much for your code snippet !
> How do I use the COO constructor to build the
> matrix from rows, cols and vals ?

A = coo_matrix( (vals, (rows, cols)) )

r.


From cscheit at lstm.uni-erlangen.de  Thu Dec 13 11:56:15 2007
From: cscheit at lstm.uni-erlangen.de (Christoph Scheit)
Date: Thu, 13 Dec 2007 17:56:15 +0100
Subject: [SciPy-user] read ascii data
Message-ID: <200712131756.15777.cscheit@lstm.uni-erlangen.de>

Thank you all,

I tried the proposed possibilities, but I guess actually that I will
go for numpy.fromfile and will switch to binary format like proposed
by Roman.
-- 

============================
M.Sc. Christoph Scheit
Institute of Fluid Mechanics
FAU Erlangen-Nuremberg
Cauerstrasse 4
D-91058 Erlangen
Phone: +49 9131 85 29508
============================


From nwagner at iam.uni-stuttgart.de  Thu Dec 13 12:07:17 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 13 Dec 2007 18:07:17 +0100
Subject: [SciPy-user] Creating a sparse matrix
In-Reply-To: <476153A7.1020707@ntc.zcu.cz>
References: <web-93277839@uni-stuttgart.de> <47612D92.4020408@ntc.zcu.cz>
	<web-93278688@uni-stuttgart.de> <47613329.4020603@ntc.zcu.cz>
	<web-93282502@uni-stuttgart.de> <476153A7.1020707@ntc.zcu.cz>
Message-ID: <web-93290641@uni-stuttgart.de>

On Thu, 13 Dec 2007 16:45:43 +0100
  Robert Cimrman <cimrman3 at ntc.zcu.cz> wrote:
> Nils Wagner wrote:
>> Thank you very much for your code snippet !
>> How do I use the COO constructor to build the
>> matrix from rows, cols and vals ?
> 
> A = coo_matrix( (vals, (rows, cols)) )
> 

Many thanks.
The file contains only the upper triangular part of the
symmetric matrix. How can I add the lower part to A ?

Nils

  


From clintonwallen at gmail.com  Thu Dec 13 12:49:39 2007
From: clintonwallen at gmail.com (Clinton Allen)
Date: Thu, 13 Dec 2007 09:49:39 -0800
Subject: [SciPy-user] Green's function
In-Reply-To: <447725.19101.qm@web34409.mail.mud.yahoo.com>
References: <c2284ea70712110304v2f7d2dcfo89504c97666b6d7d@mail.gmail.com>
	<447725.19101.qm@web34409.mail.mud.yahoo.com>
Message-ID: <c5ad77690712130949n31165402n4479e3113dbf7fdd@mail.gmail.com>

Try wikapedia: http://en.wikipedia.org/wiki/Greens_function
--Clint Allen


On Dec 11, 2007 8:52 AM, Lou Pecora <lou_boog2000 at yahoo.com> wrote:

> There is no particular special function called the
> Green Function (sometimes named Green's Function).
> Green functions are special solutions of certain
> partial differential equations (PDE), i.e. they are a
> certain class of functions.  It's like saying
> Orthogonal Polynomials -- that's a class, not a
> particular type.
>
> Green functions have the property that when used in
> the PDE they yield a Dirac delta function.  That makes
> them useful for the solution of the same PDE, but with
> a nonhomogeneous term added to the "other side."
>
> So each PDE has it's own Green function.  Eg. the
> Helmholtz equation,
>
>  D^2 G + k^2 G=0  (D^2 = Laplacian)
>
> has the Green function  G(r,r')= i H_0(k|r'-r|)/4,
> where H_0 is a Hankel function and i=sqrt(-1).
>
> I suggest you look at some texts (maybe mathematical
> physics books) for the subject.  Even a Google search
> will give you some information.
>
> -- Lou Pecora
>
> --- Natali Melgarejo Diaz <nmelgarejodiaz at gmail.com>
> wrote:
>
> > Hi to everyone i'm new in this list...i have a
> > question about a
> > formula i haven't found yet..., do you know the
> > "Green function"??. Is
> > there anything similar in Scipy??, 'cause in Matllab
> > there is one but
> > i haven't found it in Scipy yet.
> >
> > Thanks in advance to everyone ;))
> >
> >
> > ********Natali********
>
>
>
>
>  ____________________________________________________________________________________
> Never miss a thing.  Make Yahoo your home page.
> http://www.yahoo.com/r/hs
> _______________________________________________
> 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/20071213/9e3b76f6/attachment.html>

From dwf at cs.toronto.edu  Thu Dec 13 18:15:29 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 13 Dec 2007 18:15:29 -0500
Subject: [SciPy-user] arbitrary submatrices of a sparse matrix?
Message-ID: <C9311940-A4B9-4183-9F7C-1647DF7576C3@cs.toronto.edu>

Hi there,

I'm porting some MATLAB code that makes use of the Matlab convention  
of multiple index vectors to pull out an arbitrary submatrix, i.e.

x([4,6,2,7],[1,7,3])  would give you a 4x3 matrix containing rows  
4,5,6,7 and cols 1,2,3 of the original matrix.

I've figured out how to do this for full matrices (either x 
[[4,6,2,7],:][:,1,7,3]  or by playing with take()) but I'm at a loss  
for how to do it with a sparse matrix, if it's even possible (not  
sure how matlab pulls it off).

Any ideas?

Thanks,

David


From roger.herikstad at gmail.com  Thu Dec 13 19:17:36 2007
From: roger.herikstad at gmail.com (Roger Herikstad)
Date: Fri, 14 Dec 2007 08:17:36 +0800
Subject: [SciPy-user] scipy.weave.test() failures
Message-ID: <bae941750712131617s4ff74c27ncb96300a95b3f011@mail.gmail.com>

Hi all, I just built and installed the latest scipy release, but when I run
scipy.weave.test() I get these results:

>>> scipy.weave.test()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'weave'
>>> import scipy.weave
>>> scipy.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: /opt/home/roger/.python25_compiled/m7
  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
Warning: FAILURE importing tests for <module 'scipy.weave.wx_spec' from
'...s/maci/scipy/weave/wx_spec.pyc'>
/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_wx_spec.py:16:
ImportError: No module named wxPython (in <module>)
  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 '..' 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
...warning: specified build_dir '..' does not exist or is not writable.
Trying default locations
..EE.........E..........................................F..F.............................................................
======================================================================
ERROR: check_add_function_ordered (
scipy.weave.tests.test_catalog.test_catalog)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_catalog.py",
line 282, in check_add_function_ordered
    q.add_function('f',string.upper)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 649, in add_function
    self.add_function_persistent(code,function)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 665, in add_function_persistent
    cat = get_catalog(cat_dir,mode)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 300, in get_catalog
    sh = shelve.open(catalog_file,mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/shelve.py",
line 225, in open
    return DbfilenameShelf(filename, flag, protocol, writeback)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/shelve.py",
line 209, in __init__
    Shelf.__init__(self, anydbm.open(filename, flag), protocol, writeback)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/anydbm.py",
line 83, in open
    return mod.open(file, flag, mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/dbhash.py",
line 16, in open
    return bsddb.hashopen(file, flag, mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/bsddb/__init__.py",
line 306, in hashopen
    d.open(file, db.DB_HASH, flags, mode)
DBFileExistsError: (17, 'File exists -- __fop_file_setup:  Retry limit (100)
exceeded')

======================================================================
ERROR: Test persisting a function in the default catalog
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_catalog.py",
line 270, in check_add_function_persistent1
    q.add_function_persistent('code',i)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 665, in add_function_persistent
    cat = get_catalog(cat_dir,mode)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 300, in get_catalog
    sh = shelve.open(catalog_file,mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/shelve.py",
line 225, in open
    return DbfilenameShelf(filename, flag, protocol, writeback)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/shelve.py",
line 209, in __init__
    Shelf.__init__(self, anydbm.open(filename, flag), protocol, writeback)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/anydbm.py",
line 83, in open
    return mod.open(file, flag, mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/dbhash.py",
line 16, in open
    return bsddb.hashopen(file, flag, mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/bsddb/__init__.py",
line 306, in hashopen
    d.open(file, db.DB_HASH, flags, mode)
DBFileExistsError: (17, 'File exists -- __fop_file_setup:  Retry limit (100)
exceeded')

======================================================================
ERROR: Shouldn't get a single file from the temp dir.
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_catalog.py",
line 198, in check_get_existing_files2
    q.add_function('code', os.getpid)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 649, in add_function
    self.add_function_persistent(code,function)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 665, in add_function_persistent
    cat = get_catalog(cat_dir,mode)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/catalog.py",
line 300, in get_catalog
    sh = shelve.open(catalog_file,mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/shelve.py",
line 225, in open
    return DbfilenameShelf(filename, flag, protocol, writeback)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/shelve.py",
line 209, in __init__
    Shelf.__init__(self, anydbm.open(filename, flag), protocol, writeback)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/anydbm.py",
line 83, in open
    return mod.open(file, flag, mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/dbhash.py",
line 16, in open
    return bsddb.hashopen(file, flag, mode)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/bsddb/__init__.py",
line 306, in hashopen
    d.open(file, db.DB_HASH, flags, mode)
DBFileExistsError: (17, 'File exists -- __fop_file_setup:  Retry limit (100)
exceeded')

======================================================================
FAIL: check_1d_3 (
scipy.weave.tests.test_size_check.test_dummy_array_indexing)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 168, in check_1d_3
    self.generic_1d('a[-11:]')
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 135, in generic_1d
    self.generic_wrap(a,expr)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 127, in generic_wrap
    self.generic_test(a,expr,desired)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 123, in generic_test
    assert_array_equal(actual,desired, expr)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/numpy/testing/utils.py",
line 223, in assert_array_equal
    verbose=verbose, header='Arrays are not equal')
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/numpy/testing/utils.py",
line 215, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not equal
a[-11:]
(mismatch 100.0%)
 x: array([1])
 y: array([10])

======================================================================
FAIL: check_1d_6 (
scipy.weave.tests.test_size_check.test_dummy_array_indexing)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 174, in check_1d_6
    self.generic_1d('a[:-11]')
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 135, in generic_1d
    self.generic_wrap(a,expr)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 127, in generic_wrap
    self.generic_test(a,expr,desired)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/scipy/weave/tests/test_size_check.py",
line 123, in generic_test
    assert_array_equal(actual,desired, expr)
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/numpy/testing/utils.py",
line 223, in assert_array_equal
    verbose=verbose, header='Arrays are not equal')
  File
"/opt/cluster/usr/lib/python2.5/site-packages/maci/numpy/testing/utils.py",
line 215, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not equal
a[:-11]
(mismatch 100.0%)
 x: array([9])
 y: array([0])

----------------------------------------------------------------------
Ran 132 tests in 7.907s

FAILED (failures=2, errors=3)
<unittest.TextTestRunner object at 0x1cdb5d0>

This was run on an Intel Mac Pro running 10.5.1 . If I do the same on an
Intel MacBook, running the same OS version, the test passes, while still
giving the warnings about the build_dir. Any ideas as to what could be the
problem? Thanks!

~ Roger
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071214/88430dbd/attachment.html>

From dwf at cs.toronto.edu  Thu Dec 13 20:23:50 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 13 Dec 2007 20:23:50 -0500
Subject: [SciPy-user] arbitrary submatrices of a sparse matrix?
In-Reply-To: <C9311940-A4B9-4183-9F7C-1647DF7576C3@cs.toronto.edu>
References: <C9311940-A4B9-4183-9F7C-1647DF7576C3@cs.toronto.edu>
Message-ID: <9406141D-E419-4B6C-A4B4-033ACC2E1DF2@cs.toronto.edu>

On 13-Dec-07, at 6:15 PM, David Warde-Farley wrote:

> Hi there,
>
> I'm porting some MATLAB code that makes use of the Matlab convention
> of multiple index vectors to pull out an arbitrary submatrix, i.e.
>
> x([4,6,2,7],[1,7,3])  would give you a 4x3 matrix containing rows
> 4,5,6,7 and cols 1,2,3 of the original matrix.
>
> I've figured out how to do this for full matrices (either x
> [[4,6,2,7],:][:,1,7,3]  or by playing with take()) but I'm at a loss
> for how to do it with a sparse matrix, if it's even possible (not
> sure how matlab pulls it off).

Hmm, I've just noticed that get_submatrix() in SVN *claims* to do  
what I want it to do, but actually doesn't. It works fine with  
slices, but with tuples it always returns a 1x1. Is it supposed to do  
that? (There's no test case for this particular functionality and the  
documentation is, shall we say, sparse)

If there's some easy way to permute the rows and columns of a CSR/CSC  
matrix then I would be able to get the job done with slices... Is  
there? Or is there some way to do this that I'm not seeing?

Thanks again,

David


From odonnems at yahoo.com  Fri Dec 14 01:13:45 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Thu, 13 Dec 2007 22:13:45 -0800 (PST)
Subject: [SciPy-user] weave inline returning a multimap to a dict nested
	within c++
Message-ID: <665274.38999.qm@web58013.mail.re3.yahoo.com>

I would like to convert a multimap created within c++ to a python dictionary in order to return the value. I think I need to do it this way because I have a multimap (potential for redundant keys) that needs to be sorted. I am using scipy.weave.inline with converters.blitz. If you think there is a better way I am open to what ever may work. Although, I prefer to stick with weave.inline blitz converter--This is all new and there are a lot of components to wrapping c++ in python.

Anyhow, I tired this:
py::dict ckmap_freq = map_freq;
return_val = ckmap_freq;

And received this error:
error: conversion from `std::multimap<double, std::string, std::less<double>, std::allocator<std::pair<const double, std::string> > >' to non-scalar type `py::dict' requested

I was able to successfully use a similar method but with a vector instead. Therefore, it is likely I need to change the syntax.

thanks for any suggestions and help,
Michael





      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071213/1b9810e3/attachment.html>

From matthieu.brucher at gmail.com  Fri Dec 14 01:59:33 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 14 Dec 2007 07:59:33 +0100
Subject: [SciPy-user] weave inline returning a multimap to a dict nested
	within c++
In-Reply-To: <665274.38999.qm@web58013.mail.re3.yahoo.com>
References: <665274.38999.qm@web58013.mail.re3.yahoo.com>
Message-ID: <e76aa17f0712132259g6cf7acf4id038f05070169628@mail.gmail.com>

Hi,

I don't think you can transform a multimap into a dict : a multimap can have
several times the same key, a dict cannot. So first you might need to solve
this problem.

Matthieu

2007/12/14, Michael ODonnell <odonnems at yahoo.com>:
>
> I would like to convert a multimap created within c++ to a python
> dictionary in order to return the value. I think I need to do it this way
> because I have a multimap (potential for redundant keys) that needs to be
> sorted. I am using scipy.weave.inline with converters.blitz. If you think
> there is a better way I am open to what ever may work. Although, I prefer to
> stick with weave.inline blitz converter--This is all new and there are a
> lot of components to wrapping c++ in python.
>
> Anyhow, I tired this:
> py::dict ckmap_freq = map_freq;
> return_val = ckmap_freq;
>
> And received this error:
> error: conversion from `std::multimap<double, std::string,
> std::less<double>, std::allocator<std::pair<const double, std::string> > >'
> to non-scalar type `py::dict' requested
>
> I was able to successfully use a similar method but with a vector instead.
> Therefore, it is likely I need to change the syntax.
>
> thanks for any suggestions and help,
> Michael
>
>
> ------------------------------
> Never miss a thing. Make Yahoo your homepage.<http://us.rd.yahoo.com/evt=51438/*http://www.yahoo.com/r/hs>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


-- 
French PhD student
Website : http://matthieu-brucher.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/20071214/d6c623b2/attachment.html>

From nwagner at iam.uni-stuttgart.de  Fri Dec 14 03:33:41 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 14 Dec 2007 09:33:41 +0100
Subject: [SciPy-user] Rutherford Boeing format
In-Reply-To: <8793ae6e0712130703k4e6f2a87qcab343a67841748b@mail.gmail.com>
References: <web-93275518@uni-stuttgart.de>
	<8793ae6e0712130703k4e6f2a87qcab343a67841748b@mail.gmail.com>
Message-ID: <web-93319665@uni-stuttgart.de>

On Thu, 13 Dec 2007 10:03:58 -0500
  "Dominique Orban" <dominique.orban at gmail.com> wrote:
> On 12/13/07, Nils Wagner <nwagner at iam.uni-stuttgart.de> 
>wrote:
>> Hi all,
>>
>> I was wondering if someone has written a function to
>> import matrices given in the Rutherford Boeing format.
> 
> A while ago I wrote a function to import Harwell-Boeing 
>format. RB is
> quite similar so I expect few changed will do the trick. 
>My function
> relies on Konrad Hinsen's FortranFormat module, which I 
>also attach
> below. I haven't looked at this lately, so I hope it 
>works for you. 

Thank you very much for your reply. I will try it asap.

If
> this is useful, I have no problem including it in SciPy.

+1
> 

Best wishes
                   Nils

> In the test function, the sparsetools module is part of 
>NLPy and plots
> a sparsity pattern. If you don't use NLPy, just replace 
>that with any
> other plot you like :)
> 
> Dominique
  


From cimrman3 at ntc.zcu.cz  Fri Dec 14 05:21:05 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Fri, 14 Dec 2007 11:21:05 +0100
Subject: [SciPy-user] arbitrary submatrices of a sparse matrix?
In-Reply-To: <9406141D-E419-4B6C-A4B4-033ACC2E1DF2@cs.toronto.edu>
References: <C9311940-A4B9-4183-9F7C-1647DF7576C3@cs.toronto.edu>
	<9406141D-E419-4B6C-A4B4-033ACC2E1DF2@cs.toronto.edu>
Message-ID: <47625911.1010400@ntc.zcu.cz>

David Warde-Farley wrote:
> Hmm, I've just noticed that get_submatrix() in SVN *claims* to do  
> what I want it to do, but actually doesn't. It works fine with  
> slices, but with tuples it always returns a 1x1. Is it supposed to do  
> that? (There's no test case for this particular functionality and the  
> documentation is, shall we say, sparse)

It works for slices only. Yes, the docstring, as I reread it now, is not 
clear. What it should say is, that the slice can be given as:
1. slice object
2. a tuple (from, to)
3. a scalar for single row/column selection
The method should probably be named _get_submatrix(), as it is used 
internally in __getitem__().

r.


From cimrman3 at ntc.zcu.cz  Fri Dec 14 10:00:42 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Fri, 14 Dec 2007 16:00:42 +0100
Subject: [SciPy-user] ANN: SFE-00.35.01
Message-ID: <47629A9A.8030708@ntc.zcu.cz>

Let me announce SFE-00.35.01, bringing per term integration - now each 
term can use its own quadrature points. This is a major change at the 
heart of the code - some parts may not work as all terms were not 
migrated yet to the new framework. All test examples work, though, as 
well as acoustic band gaps. See http://ui505p06-mbs.ntc.zcu.cz/sfe .

SFE is a finite element analysis software written almost entirely in 
Python. The code is released under BSD license.

best regards,
r.


From s.mientki at ru.nl  Fri Dec 14 10:52:44 2007
From: s.mientki at ru.nl (Stef Mientki)
Date: Fri, 14 Dec 2007 16:52:44 +0100
Subject: [SciPy-user] static magnetic field, simulation ?
Message-ID: <4762A6CC.80602@ru.nl>

hello,

Does anyone knows an implementation of the calculation and vizialization
of a static magnetic field and the induced forces ?
I couldn't find anything in python on this subject,
and outside python there's also almost nothing to find.

Background, my son is doing a school project on magnetic levitation,
my knowledge about magnetic fields is too rusty,
and it would be a very nice example to put in PyLab Works ;-)

thanks,
Stef Mientki


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 bryanv at enthought.com  Fri Dec 14 11:16:32 2007
From: bryanv at enthought.com (Bryan Van de Ven)
Date: Fri, 14 Dec 2007 10:16:32 -0600
Subject: [SciPy-user] static magnetic field, simulation ?
In-Reply-To: <4762A6CC.80602@ru.nl>
References: <4762A6CC.80602@ru.nl>
Message-ID: <4762AC60.2060808@enthought.com>

http://www.mare.ee/indrek/ephi/ might have what you want.


Stef Mientki wrote:
> hello,
> 
> Does anyone knows an implementation of the calculation and vizialization
> of a static magnetic field and the induced forces ?
> I couldn't find anything in python on this subject,
> and outside python there's also almost nothing to find.
> 
> Background, my son is doing a school project on magnetic levitation,
> my knowledge about magnetic fields is too rusty,
> and it would be a very nice example to put in PyLab Works ;-)
> 
> thanks,
> Stef Mientki
> 
> 
> 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.
> 
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
> 



From nwagner at iam.uni-stuttgart.de  Fri Dec 14 14:04:39 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 14 Dec 2007 20:04:39 +0100
Subject: [SciPy-user] Coo_constructor
Message-ID: <web-93353981@uni-stuttgart.de>

Hi all,

I use the coo constructor to build sparse matrices.
A_i = coo_matrix( (vals_i, (rows_i, cols_i)) )

I read the data for each matrix from a different file.

One of the matrices is rank-one with only non-zero entry.

(7, 7)        5.0

Another matrix has the entries

   (0, 0)        80000000.0
   (0, 1)        -120000000.0
   (1, 0)        -120000000.0
   (1, 1)        480000000.0
   (0, 2)        40000000.0
   (2, 0)        40000000.0
   (2, 2)        160000000.0
   (1, 3)        -240000000.0
   (3, 1)        -240000000.0
   (2, 3)        -120000000.0
   (3, 2)        -120000000.0
   (3, 3)        480000000.0
   (1, 4)        120000000.0
   (4, 1)        120000000.0
   (2, 4)        40000000.0
   (4, 2)        40000000.0
   (4, 4)        160000000.0
   (3, 5)        -240000000.0
   (5, 3)        -240000000.0
   (4, 5)        -120000000.0
   (5, 4)        -120000000.0
   (5, 5)        480000000.0
   (3, 6)        120000000.0
   (6, 3)        120000000.0
   (4, 6)        40000000.0
   :     :
   (9, 11)       -240000000.0
   (11, 9)       -240000000.0
   (10, 11)      -120000000.0
   (11, 10)      -120000000.0
   (11, 11)      480000000.0
   (9, 12)       120000000.0
   (12, 9)       120000000.0
   (10, 12)      40000000.0
   (12, 10)      40000000.0
   (12, 12)      160000000.0
   (11, 13)      -240000000.0
   (13, 11)      -240000000.0
   (12, 13)      -120000000.0
   (13, 12)      -120000000.0
   (13, 13)      480000000.0
   (11, 14)      120000000.0
   (14, 11)      120000000.0
   (12, 14)      40000000.0
   (14, 12)      40000000.0
   (14, 14)      160000000.0
   (13, 15)      120000000.0
   (15, 13)      120000000.0
   (14, 15)      40000000.0
   (15, 14)      40000000.0
   (15, 15)      80000000.0


How can I circumvent the
ValueError: shape mismatch: objects cannot be broadcast to 
a single shape

if I try to add the matrices ?

Nils




From lbolla at gmail.com  Fri Dec 14 14:31:22 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Fri, 14 Dec 2007 20:31:22 +0100
Subject: [SciPy-user] Coo_constructor
In-Reply-To: <web-93353981@uni-stuttgart.de>
References: <web-93353981@uni-stuttgart.de>
Message-ID: <80c99e790712141131o3d8ee818x730f4f5e44d8d4d2@mail.gmail.com>

You can always define the sparse matrix dimension in this way:
scipy.sparse.coo_matrix(([5.], [[7.],[7.]]), dims=(16, 16))

L.

On Dec 14, 2007 8:04 PM, Nils Wagner <nwagner at iam.uni-stuttgart.de> wrote:

> Hi all,
>
> I use the coo constructor to build sparse matrices.
> A_i = coo_matrix( (vals_i, (rows_i, cols_i)) )
>
> I read the data for each matrix from a different file.
>
> One of the matrices is rank-one with only non-zero entry.
>
> (7, 7)        5.0
>
> Another matrix has the entries
>
>   (0, 0)        80000000.0
>   (0, 1)        -120000000.0
>   (1, 0)        -120000000.0
>   (1, 1)        480000000.0
>   (0, 2)        40000000.0
>   (2, 0)        40000000.0
>   (2, 2)        160000000.0
>   (1, 3)        -240000000.0
>   (3, 1)        -240000000.0
>   (2, 3)        -120000000.0
>   (3, 2)        -120000000.0
>   (3, 3)        480000000.0
>   (1, 4)        120000000.0
>   (4, 1)        120000000.0
>   (2, 4)        40000000.0
>   (4, 2)        40000000.0
>   (4, 4)        160000000.0
>   (3, 5)        -240000000.0
>   (5, 3)        -240000000.0
>   (4, 5)        -120000000.0
>   (5, 4)        -120000000.0
>   (5, 5)        480000000.0
>   (3, 6)        120000000.0
>   (6, 3)        120000000.0
>   (4, 6)        40000000.0
>   :     :
>   (9, 11)       -240000000.0
>   (11, 9)       -240000000.0
>   (10, 11)      -120000000.0
>   (11, 10)      -120000000.0
>   (11, 11)      480000000.0
>   (9, 12)       120000000.0
>   (12, 9)       120000000.0
>   (10, 12)      40000000.0
>   (12, 10)      40000000.0
>   (12, 12)      160000000.0
>   (11, 13)      -240000000.0
>   (13, 11)      -240000000.0
>   (12, 13)      -120000000.0
>   (13, 12)      -120000000.0
>   (13, 13)      480000000.0
>   (11, 14)      120000000.0
>   (14, 11)      120000000.0
>   (12, 14)      40000000.0
>   (14, 12)      40000000.0
>   (14, 14)      160000000.0
>   (13, 15)      120000000.0
>   (15, 13)      120000000.0
>   (14, 15)      40000000.0
>   (15, 14)      40000000.0
>   (15, 15)      80000000.0
>
>
> How can I circumvent the
> ValueError: shape mismatch: objects cannot be broadcast to
> a single shape
>
> if I try to add the matrices ?
>
> Nils
>
>
> _______________________________________________
> 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/20071214/45da190a/attachment.html>

From dwf at cs.toronto.edu  Fri Dec 14 19:12:19 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Fri, 14 Dec 2007 19:12:19 -0500
Subject: [SciPy-user] Sparse with fast element-wise multiply?
Message-ID: <27261D40-C6F1-402B-AFE3-D6B333CDA4FB@cs.toronto.edu>

Hi folks,

I'm currently using sparse.lil_matrix to do element-wise  
multiplication of two sparse matrices, and, well, it's really slow.  
Any suggestions for something speedier? (it seems csr and csc matrices  
don't have .multiply()).

David




From dwf at cs.toronto.edu  Fri Dec 14 19:21:43 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Fri, 14 Dec 2007 19:21:43 -0500
Subject: [SciPy-user] arbitrary submatrices of a sparse matrix?
In-Reply-To: <47625911.1010400@ntc.zcu.cz>
References: <C9311940-A4B9-4183-9F7C-1647DF7576C3@cs.toronto.edu>
	<9406141D-E419-4B6C-A4B4-033ACC2E1DF2@cs.toronto.edu>
	<47625911.1010400@ntc.zcu.cz>
Message-ID: <C39ED495-3493-44FC-A93E-9A90BBC6EC71@cs.toronto.edu>

On 14-Dec-07, at 5:21 AM, Robert Cimrman wrote:

> It works for slices only. Yes, the docstring, as I reread it now, is  
> not
> clear. What it should say is, that the slice can be given as:
> 1. slice object
> 2. a tuple (from, to)
> 3. a scalar for single row/column selection
> The method should probably be named _get_submatrix(), as it is used
> internally in __getitem__().

Ah, I figured that out, yes, the tuple is just a start/end point.  I  
ended up using the coo_matrix format and just manually do the element  
selection and reindexing. I've pasted a copy of my code here, so that  
it's in the archives (and so that others may comment and suggest  
improvements).

I'll also submit a patch of sparse.py that adds a method for this, in  
case the authors think it warrants inclusion.

Cheers,

DWF



def coo_submatrix_pull(matr, rows, cols):
     """
     Pulls out an arbitrary i.e. non-contiguous submatrix out of
     a sparse.coo_matrix. This does create a copy, however, which
     is probably not ideal.
     """
     if type(matr) != S.sparse.coo_matrix:
         raise TypeError('Matrix must be sparse COOrdinate format')

     gr = -1 * ones(matr.shape[0])
     gc = -1 * ones(matr.shape[1])

     lr = len(rows)
     lc = len(cols)

     ar = arange(0, lr)
     ac = arange(0, lc)
     gr[rows[ar]] = ar
     gc[cols[ac]] = ac

     newelem = (gr[matr.row] > -1) & (gc[matr.col] > -1)
     newrows = matr.row[newelem]
     newcols = matr.col[newelem]

     return S.sparse.coo_matrix((matr.data[newelem],  
array([gr[newrows], \
         gc[newcols]])),(lr, lc))


From tjhnson at gmail.com  Sat Dec 15 00:26:59 2007
From: tjhnson at gmail.com (Tom Johnson)
Date: Fri, 14 Dec 2007 21:26:59 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
Message-ID: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>

While playing with Mathematica, I must say that I was surprised to
find that it handled element-wise multiplication differently from
scipy.

In scipy,

>>> A = array([[1,2],[3,4]])
>>> B = array([2,3])
>>> A * B
array([[2,6],[6,12]])

which essentially multiplies each COLUMN of A with each COLUMN of B.


In mathematica,

>>> A = {{1,2},{3,4}}
>>> B = {2,3}
>>> A * B
{{2,4},{9,12}}

which essentially multiples each ROW of A with each COLUMN of B.  I
actually find this method of multiplication more intuitive (as it is
like matrix multiplication).

Now, scipy can easily obtain the Mathematica result by the following:

>>> A = array([[1,2],[3,4]])
>>> B = array([[2],[3]])
>>> A * B
array([[2,4],[9,12]])

but it isn't as easy to get Mathematica to return the scipy
result...that is, the following natural attempt does not work:

>>> A = {{1,2},{3,4}}
>>> B = {{2},{3}}
>>> A * B
Thread::tdlen: Objects of unequal length in {1,2}{2}  cannot be combined.
Thread::tdlen: Objects of unequal length in {3,4}{3}  cannot be combined.

and it is easy to see why Mathematica is unable to complete this
operation...but of course there are other ways to achieve this result.


Anyway, I just thought the difference was interesting....


From stefan at sun.ac.za  Sat Dec 15 06:31:42 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Sat, 15 Dec 2007 13:31:42 +0200
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>
Message-ID: <20071215113142.GH7167@mentat.za.net>

Hi Tom

On Fri, Dec 14, 2007 at 09:26:59PM -0800, Tom Johnson wrote:
> While playing with Mathematica, I must say that I was surprised to
> find that it handled element-wise multiplication differently from
> scipy.
> 
> In scipy,
> 
> >>> A = array([[1,2],[3,4]])
> >>> B = array([2,3])
> >>> A * B
> array([[2,6],[6,12]])
> 
> which essentially multiplies each COLUMN of A with each COLUMN of B.

Numpy does what we call broadcasting.  You can read more about it at

http://www.scipy.org/EricsBroadcastingDoc

If you'd like to have the Mathematica behaviour, you can always use
the transpose property (i.e. x.T).

Regards
St?fan


From rex at nosyntax.net  Sat Dec 15 12:37:22 2007
From: rex at nosyntax.net (rex)
Date: Sat, 15 Dec 2007 09:37:22 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <20071215113142.GH7167@mentat.za.net>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>
	<20071215113142.GH7167@mentat.za.net>
Message-ID: <20071215173722.GF24062@nosyntax.net>

Stefan van der Walt <stefan at sun.ac.za> [2007-12-15 03:32]:
>Numpy does what we call broadcasting.  You can read more about it at
>
>http://www.scipy.org/EricsBroadcastingDoc

Very nice! A big "Thank you!" to Eric for making such clear
documentation available. The diagrams really help.

-rex


From dmitrey.kroshko at scipy.org  Sat Dec 15 15:17:06 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Sat, 15 Dec 2007 22:17:06 +0200
Subject: [SciPy-user] Ann: OpenOpt v 0.15 (free optimization framework)
Message-ID: <47643642.80808@scipy.org>

Hi all,
we are glad to inform you about OpenOpt v 0.15 (release), free (license: 
BSD) optimization framework for Python language programmers

Changes since previous release (September 4):

    * some new classes
    * several new solvers written
    * some more solvers connected
    * NLP/NSP solver ralg can handle constrained problems
    * some bugfixes
    * some enhancements in graphical output (especially for constrained 
problems)

Regards,
OpenOpt developers
http://scipy.org/scipy/scikits/wiki/OpenOpt
http://openopt.blogspot.com/


From lbolla at gmail.com  Sun Dec 16 14:03:49 2007
From: lbolla at gmail.com (lorenzo bolla)
Date: Sun, 16 Dec 2007 20:03:49 +0100
Subject: [SciPy-user] Sparse with fast element-wise multiply?
In-Reply-To: <27261D40-C6F1-402B-AFE3-D6B333CDA4FB@cs.toronto.edu>
References: <27261D40-C6F1-402B-AFE3-D6B333CDA4FB@cs.toronto.edu>
Message-ID: <80c99e790712161103j4f5ae3ecj4b53d6100403dea8@mail.gmail.com>

converting the matrix from lil_matrix to csc or csr format gives a 140x
speed improvement on multiplication.

In [4]: A = scipy.sparse.lil_matrix((1000, 1000))

In [5]: A[:,100] = scipy.rand(1000)

In [6]: A[99, :] = scipy.rand(1000)

In [7]: A.setdiag(scipy.rand(1000))

In [8]: %timeit A * A
10 loops, best of 3: 72.2 ms per loop

In [9]: B = A.tocsr()

In [10]: %timeit B * B
1000 loops, best of 3: 510 ?s per loop

In [11]: C = A.tocsc()

In [12]: %timeit C * C
1000 loops, best of 3: 527 ?s per loop

hth,
L.





On Dec 15, 2007 1:12 AM, David Warde-Farley <dwf at cs.toronto.edu> wrote:

> Hi folks,
>
> I'm currently using sparse.lil_matrix to do element-wise
> multiplication of two sparse matrices, and, well, it's really slow.
> Any suggestions for something speedier? (it seems csr and csc matrices
> don't have .multiply()).
>
> 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/20071216/570c2b6f/attachment.html>

From dwf at cs.toronto.edu  Sun Dec 16 14:50:12 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Sun, 16 Dec 2007 14:50:12 -0500
Subject: [SciPy-user] Sparse with fast element-wise multiply?
In-Reply-To: <80c99e790712161103j4f5ae3ecj4b53d6100403dea8@mail.gmail.com>
References: <27261D40-C6F1-402B-AFE3-D6B333CDA4FB@cs.toronto.edu>
	<80c99e790712161103j4f5ae3ecj4b53d6100403dea8@mail.gmail.com>
Message-ID: <65C83323-C9E0-487A-8CFE-A5238B0A4C57@cs.toronto.edu>

On 16-Dec-07, at 2:03 PM, lorenzo bolla wrote:

> converting the matrix from lil_matrix to csc or csr format gives a  
> 140x speed improvement on multiplication.
>
> In [4]: A = scipy.sparse.lil_matrix((1000, 1000))
>
> In [5]: A[:,100] = scipy.rand(1000)
>
> In [6]: A[99, :] = scipy.rand(1000)
>
> In [7]: A.setdiag(scipy.rand(1000))

Hi Lorenzo,

Thanks for your reply. Actually, though, I'm not looking for matrix- 
multiply; I know CSR/CSC is faster for that.

I'm looking for elementwise multiply, i.e. take two matrices of the  
same size, multiply each element in matrix 1 with the corresponding  
element in matrix 2, and put the result in the same position in the  
result matrix (as with matlab's a .* b syntax). It seems lil_matrix is  
the only sparse type that implements .multiply(), and converting to  
lil_matrix is slow, so I've written my own function for multiplying  
two coo_matrix's together. It's a LITTLE slower than .multiply() but  
faster than converting both matrices to lil_matrix and then multiplying.

I'm not entirely sure how CSR and CSC work so it might be possible to  
implement faster elementwise multiplication on top of them.

David



From cohen at slac.stanford.edu  Sun Dec 16 19:28:14 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Sun, 16 Dec 2007 16:28:14 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <20071215113142.GH7167@mentat.za.net>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>
	<20071215113142.GH7167@mentat.za.net>
Message-ID: <4765C29E.6070306@slac.stanford.edu>

Actually I do not manage to use .T or .transpose() method on 1D arrays :

In [42]: a = array([[ 0.0, 0.0, 
0.0],[10.0,10.0,10.0],[20.0,20.0,20.0],[30.0,30.0,30.0]])   <--this is 
example 3 of thiis indeed very nice tutorial on broadcasting
In [43]: b = array([1.0,2.0,3.0,4.0])
In [44]: a+b
ValueError: shape mismatch: objects cannot be broadcast to a single 
shape    <----- fine, mismatch of trailing dimensions

In [45]: b.transpose()
Out[45]: array([ 1.,  2.,  3.,  4.])     <------ ominous : no change

In [47]: a+b.transpose()
ValueError: shape mismatch: objects cannot be broadcast to a single 
shape    <-----transpose did not work

In [48]: a+b.T
ValueError: shape mismatch: objects cannot be broadcast to a single 
shape    <------ nor .T

In [49]: b = array([[1.0],[2.0],[3.0],[4.0]])
In [51]: a+b
Out[51]:
array([[  1.,   1.,   1.],
       [ 12.,  12.,  12.],
       [ 23.,  23.,  23.],
       [ 34.,  34.,  34.]])     <------------ expected result for the 
brodcast of b as a column vector into a

In [52]: aa=array([[1,2],[3,4]])

In [53]: aa.transpose()
Out[53]:
array([[1, 3],
       [2, 4]])
                                      <------------- transpose and T 
work for this 2D array
In [55]: aa.T
Out[55]:
array([[1, 3],
       [2, 4]])


So, what is going on? Is that a bug, a feature, or some weir failure of 
my install?

thanks,
Johann

Stefan van der Walt wrote:
> Hi Tom
>
> On Fri, Dec 14, 2007 at 09:26:59PM -0800, Tom Johnson wrote:
>   
>> While playing with Mathematica, I must say that I was surprised to
>> find that it handled element-wise multiplication differently from
>> scipy.
>>
>> In scipy,
>>
>>     
>>>>> A = array([[1,2],[3,4]])
>>>>> B = array([2,3])
>>>>> A * B
>>>>>           
>> array([[2,6],[6,12]])
>>
>> which essentially multiplies each COLUMN of A with each COLUMN of B.
>>     
>
> Numpy does what we call broadcasting.  You can read more about it at
>
> http://www.scipy.org/EricsBroadcastingDoc
>
> If you'd like to have the Mathematica behaviour, you can always use
> the transpose property (i.e. x.T).
>
> Regards
> St?fan
> _______________________________________________
> 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  Sun Dec 16 21:28:48 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 17 Dec 2007 11:28:48 +0900
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <4765C29E.6070306@slac.stanford.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>
	<4765C29E.6070306@slac.stanford.edu>
Message-ID: <4765DEE0.9030401@ar.media.kyoto-u.ac.jp>

Johann Cohen-Tanugi wrote:
> Actually I do not manage to use .T or .transpose() method on 1D arrays :
>
> In [42]: a = array([[ 0.0, 0.0, 
> 0.0],[10.0,10.0,10.0],[20.0,20.0,20.0],[30.0,30.0,30.0]])   <--this is 
> example 3 of thiis indeed very nice tutorial on broadcasting
> In [43]: b = array([1.0,2.0,3.0,4.0])
> In [44]: a+b
> ValueError: shape mismatch: objects cannot be broadcast to a single 
> shape    <----- fine, mismatch of trailing dimensions
>
> In [45]: b.transpose()
> Out[45]: array([ 1.,  2.,  3.,  4.])     <------ ominous : no change
This is confusing if you are coming from matlab and similar softwares 
(it was for me, at least). In matlab, any array is rank 2 by default 
(let's put aside rank > 2 for now). That is, in matlab, a = [1, 2, 3] 
gives you a row array, which is interpreted as a matrix: size(a) gives 
you [1, 3]. In numpy, this is not the case: there is a real difference 
between "row arrays", "column arrays" and vectors.

a = array([1, 2, 3])
a.ndim <----------- is 1 (rank 1)
a = array([[1], [2], [3]])
a.ndim <----------- is 2 (rank 2: column)
a = array([[1, 2, 3]])
a.ndim <----------- is 2 (rank 2: row)

In Matlab, there is no such difference, and it is really ingrained in 
the software (for example, at the C api level, the function to get the 
number of dimensions, alas the 'rank', always returns at least 2). To 
solve your problem, you should have a new dimension:

b = array([1., 2, 3])
b.ndim <----------- 1
b[:, numpy.newaxis]
b.ndim <----------- 2 (this will be a column array: b is now exactly the 
same as array([[1.], [2], [3]])
b = array([1., 2, 3])
b[numpy.newaxis, :]
b.ndim <----------- 2 (this will be a row array: b is now exactly the 
same as array([[1., 2, 3]])

cheers,

David


From cohen at slac.stanford.edu  Mon Dec 17 00:10:15 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Sun, 16 Dec 2007 21:10:15 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <4765DEE0.9030401@ar.media.kyoto-u.ac.jp>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>	<4765C29E.6070306@slac.stanford.edu>
	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>
Message-ID: <476604B7.4060400@slac.stanford.edu>

thanks for these precisions, David. Reading it, I still come to think 
that it is a potential source of confusion to let a "row array" have a 
transpose or T method, that essentially does nothing. I guess there is a 
reason behind this situation, but given the fact that it is there, I am 
wondering whether T or transpose of a row array could in fact return 
what it is expected to, aka the 2d column array. Is there any reason not 
to have this functionality?
best,
Johann

David Cournapeau wrote:
> Johann Cohen-Tanugi wrote:
>   
>> Actually I do not manage to use .T or .transpose() method on 1D arrays :
>>
>> In [42]: a = array([[ 0.0, 0.0, 
>> 0.0],[10.0,10.0,10.0],[20.0,20.0,20.0],[30.0,30.0,30.0]])   <--this is 
>> example 3 of thiis indeed very nice tutorial on broadcasting
>> In [43]: b = array([1.0,2.0,3.0,4.0])
>> In [44]: a+b
>> ValueError: shape mismatch: objects cannot be broadcast to a single 
>> shape    <----- fine, mismatch of trailing dimensions
>>
>> In [45]: b.transpose()
>> Out[45]: array([ 1.,  2.,  3.,  4.])     <------ ominous : no change
>>     
> This is confusing if you are coming from matlab and similar softwares 
> (it was for me, at least). In matlab, any array is rank 2 by default 
> (let's put aside rank > 2 for now). That is, in matlab, a = [1, 2, 3] 
> gives you a row array, which is interpreted as a matrix: size(a) gives 
> you [1, 3]. In numpy, this is not the case: there is a real difference 
> between "row arrays", "column arrays" and vectors.
>
> a = array([1, 2, 3])
> a.ndim <----------- is 1 (rank 1)
> a = array([[1], [2], [3]])
> a.ndim <----------- is 2 (rank 2: column)
> a = array([[1, 2, 3]])
> a.ndim <----------- is 2 (rank 2: row)
>
> In Matlab, there is no such difference, and it is really ingrained in 
> the software (for example, at the C api level, the function to get the 
> number of dimensions, alas the 'rank', always returns at least 2). To 
> solve your problem, you should have a new dimension:
>
> b = array([1., 2, 3])
> b.ndim <----------- 1
> b[:, numpy.newaxis]
> b.ndim <----------- 2 (this will be a column array: b is now exactly the 
> same as array([[1.], [2], [3]])
> b = array([1., 2, 3])
> b[numpy.newaxis, :]
> b.ndim <----------- 2 (this will be a row array: b is now exactly the 
> same as array([[1., 2, 3]])
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From matthieu.brucher at gmail.com  Mon Dec 17 01:48:05 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Mon, 17 Dec 2007 07:48:05 +0100
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <476604B7.4060400@slac.stanford.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>
	<20071215113142.GH7167@mentat.za.net>
	<4765C29E.6070306@slac.stanford.edu>
	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>
	<476604B7.4060400@slac.stanford.edu>
Message-ID: <e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>

2007/12/17, Johann Cohen-Tanugi <cohen at slac.stanford.edu>:
>
> thanks for these precisions, David. Reading it, I still come to think
> that it is a potential source of confusion to let a "row array" have a
> transpose or T method, that essentially does nothing.



In object oriented code, this can happen often, but it is not a problem. It
does what you want : inverse the axis, even if there is only one axis.


I guess there is a
> reason behind this situation, but given the fact that it is there, I am
> wondering whether T or transpose of a row array could in fact return
> what it is expected to, aka the 2d column array. Is there any reason not
> to have this functionality?



More code in a simple function (thus slower) ? Breaking some code that
depend on it ? Not doing what the documentation says ?
I think the method should not be changed, it does exactly what it says it
does. Changing its behaviour because of Matlab is not a good solution,
Python is not Matlab.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071217/7a458ddf/attachment.html>

From cohen at slac.stanford.edu  Mon Dec 17 02:49:38 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Sun, 16 Dec 2007 23:49:38 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>	<4765C29E.6070306@slac.stanford.edu>	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>	<476604B7.4060400@slac.stanford.edu>
	<e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
Message-ID: <47662A12.7060107@slac.stanford.edu>

Matthieu Brucher wrote:
>
>
> 2007/12/17, Johann Cohen-Tanugi <cohen at slac.stanford.edu 
> <mailto:cohen at slac.stanford.edu>>:
>
>     thanks for these precisions, David. Reading it, I still come to think
>     that it is a potential source of confusion to let a "row array" have a
>     transpose or T method, that essentially does nothing.
>
>
>
> In object oriented code, this can happen often, but it is not a 
> problem. It does what you want : inverse the axis, even if there is 
> only one axis.
hmmm...... okay... What I wanted was to transpose a 1D array into a 
vector, or vice-versa, with the linear algebra behavior in mind. I 
understand that numpy does not follow this, but I cannot believe that 
this behavior *is* what everybody wants! Tom's initial email was 
symptomatic, and Stefan's response, with the proposal to use the T 
method even more so!

Assuming that this natural linear algebra could be retrieved when, and 
*only* when, the array is 1D, I do not see how such an implementation 
could break codes that depend on it, because I don't see why someone 
would call 'a.T' just to have 'a' again.... But it is probably my lack 
of imagination.
Anyway, enough of this. I am sure the developers know better than me...

cheers,
Johann

>  
>
>     I guess there is a
>     reason behind this situation, but given the fact that it is there,
>     I am
>     wondering whether T or transpose of a row array could in fact return
>     what it is expected to, aka the 2d column array. Is there any
>     reason not
>     to have this functionality?
>
>
>
> More code in a simple function (thus slower) ? Breaking some code that 
> depend on it ? Not doing what the documentation says ?
> I think the method should not be changed, it does exactly what it says 
> it does. Changing its behaviour because of Matlab is not a good 
> solution, Python is not Matlab.
>  
> Matthieu
> -- 
> French PhD student
> Website : http://matthieu-brucher.developpez.com/
> Blogs : http://matt.eifelle.com <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
>   



From aisaac at american.edu  Mon Dec 17 08:04:01 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Mon, 17 Dec 2007 08:04:01 -0500
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <47662A12.7060107@slac.stanford.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>	<4765C29E.6070306@slac.stanford.edu>	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>	<476604B7.4060400@slac.stanford.edu><e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com><47662A12.7060107@slac.stanford.edu>
Message-ID: <Mahogany-0.67.0-1412-20071217-080401.00@american.edu>

On Sun, 16 Dec 2007, Johann Cohen-Tanugi apparently wrote:
> Assuming that this natural linear algebra could be 
> retrieved when, and only when, the array is 1D 

You appear to be talking about a "natural" linear algebra 
for 2D objects.  For example, when textbooks talk of row 
vectors and column vectors, they are talking about 2D 
objects, even though they use a simplified indexing.  (This 
vocabulary is only slightly misleading: after all, matrices 
of a given dimension are also vectors and can be indexed 
with a single index---Matlab allows this (column major) and 
NumPy allows this too (row major, via the `flat` 
attribute).)

If that is what you always seek, stick with matrices::

            >>> x=numpy.mat([0,1,2])
            >>> x
            matrix([[0, 1, 2]])
            >>> x.T
            matrix([[0],
                    [1],
                    [2]])
            >>>

In the meantime, you can always add an axis if you wish::

>>> x
array([0, 1, 2])
>>> y=x[:,None]
>>> y
array([[0],
       [1],
       [2]])

hth,
Alan Isaac





From matthieu.brucher at gmail.com  Mon Dec 17 09:01:23 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Mon, 17 Dec 2007 15:01:23 +0100
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <47662A12.7060107@slac.stanford.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>
	<20071215113142.GH7167@mentat.za.net>
	<4765C29E.6070306@slac.stanford.edu>
	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>
	<476604B7.4060400@slac.stanford.edu>
	<e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
	<47662A12.7060107@slac.stanford.edu>
Message-ID: <e76aa17f0712170601m5a9eca45kff7a20dfae002a3f@mail.gmail.com>

>
> > In object oriented code, this can happen often, but it is not a
> > problem. It does what you want : inverse the axis, even if there is
> > only one axis.
> hmmm...... okay... What I wanted was to transpose a 1D array into a
> vector, or vice-versa, with the linear algebra behavior in mind. I
> understand that numpy does not follow this, but I cannot believe that
> this behavior *is* what everybody wants! Tom's initial email was
> symptomatic, and Stefan's response, with the proposal to use the T
> method even more so!



I think that waht you really want is to use the matrix class instead of the
array. Matrix always has 2 dimensions.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071217/52cd652d/attachment.html>

From bblais at bryant.edu  Mon Dec 17 10:29:13 2007
From: bblais at bryant.edu (Brian Blais)
Date: Mon, 17 Dec 2007 10:29:13 -0500
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <47662A12.7060107@slac.stanford.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>	<4765C29E.6070306@slac.stanford.edu>	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>	<476604B7.4060400@slac.stanford.edu>
	<e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
	<47662A12.7060107@slac.stanford.edu>
Message-ID: <320E5E02-63B0-47B4-9780-B77BE578E786@bryant.edu>


On Dec 17, 2007, at Dec 17:2:49 AM, Johann Cohen-Tanugi wrote:

> Matthieu Brucher wrote:
>>
>>
>> 2007/12/17, Johann Cohen-Tanugi <cohen at slac.stanford.edu
>> <mailto:cohen at slac.stanford.edu>>:
>>
>>     thanks for these precisions, David. Reading it, I still come  
>> to think
>>     that it is a potential source of confusion to let a "row  
>> array" have a
>>     transpose or T method, that essentially does nothing.
>>
>>
>>
>> In object oriented code, this can happen often, but it is not a
>> problem. It does what you want : inverse the axis, even if there is
>> only one axis.
> hmmm...... okay... What I wanted was to transpose a 1D array into a
> vector, or vice-versa, with the linear algebra behavior in mind.

if you have linear algebra in mind, then you can use a matrix...it  
works the way you expect:

In [1]:from numpy import *

In [2]:a=array([1,2,3,4])

In [3]:a.shape
Out[3]:(4,)

In [4]:b=a.T

In [5]:b.shape
Out[5]:(4,)

In [6]:a=matrix([1,2,3,4])

In [7]:a.shape
Out[7]:(1, 4)

In [8]:b=a.T

In [9]:b.shape
Out[9]:(4, 1)

In [10]:a
Out[10]:matrix([[1, 2, 3, 4]])

In [11]:b
Out[11]:
matrix([[1],
         [2],
         [3],
         [4]])




			bb
-- 
Brian Blais
bblais at bryant.edu
http://web.bryant.edu/~bblais



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071217/7a6e6d8c/attachment.html>

From cohen at slac.stanford.edu  Mon Dec 17 11:30:20 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Mon, 17 Dec 2007 08:30:20 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <320E5E02-63B0-47B4-9780-B77BE578E786@bryant.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>	<4765C29E.6070306@slac.stanford.edu>	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>	<476604B7.4060400@slac.stanford.edu>
	<e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
	<47662A12.7060107@slac.stanford.edu>
	<320E5E02-63B0-47B4-9780-B77BE578E786@bryant.edu>
Message-ID: <4766A41C.4040200@slac.stanford.edu>

hi Brian,
point well taken. Thank you! I definitely use array more than I should then.
best,
Johann

Brian Blais wrote:
>
> On Dec 17, 2007, at Dec 17:2:49 AM, Johann Cohen-Tanugi wrote:
>
>> Matthieu Brucher wrote:
>>>
>>>
>>> 2007/12/17, Johann Cohen-Tanugi <cohen at slac.stanford.edu 
>>> <mailto:cohen at slac.stanford.edu> 
>>> <mailto:cohen at slac.stanford.edu>>:
>>>
>>>     thanks for these precisions, David. Reading it, I still come to 
>>> think
>>>     that it is a potential source of confusion to let a "row array" 
>>> have a
>>>     transpose or T method, that essentially does nothing.
>>>
>>>
>>>
>>> In object oriented code, this can happen often, but it is not a 
>>> problem. It does what you want : inverse the axis, even if there is 
>>> only one axis.
>> hmmm...... okay... What I wanted was to transpose a 1D array into a 
>> vector, or vice-versa, with the linear algebra behavior in mind.
>
> if you have linear algebra in mind, then you can use a matrix...it 
> works the way you expect:
>
> In [1]:from numpy import *
>
> In [2]:a=array([1,2,3,4])
>
> In [3]:a.shape
> Out[3]:(4,)
>
> In [4]:b=a.T
>
> In [5]:b.shape
> Out[5]:(4,)
>
> In [6]:a=matrix([1,2,3,4])
>
> In [7]:a.shape
> Out[7]:(1, 4)
>
> In [8]:b=a.T
>
> In [9]:b.shape
> Out[9]:(4, 1)
>
> In [10]:a
> Out[10]:matrix([[1, 2, 3, 4]])
>
> In [11]:b
> Out[11]:
> matrix([[1],
>         [2],
>         [3],
>         [4]])
>
>
>
>
> bb
> -- 
> Brian Blais
> bblais at bryant.edu <mailto:bblais at bryant.edu>
> http://web.bryant.edu/~bblais <http://web.bryant.edu/%7Ebblais>
>
>
>



From cscheit at lstm.uni-erlangen.de  Mon Dec 17 12:12:11 2007
From: cscheit at lstm.uni-erlangen.de (Christoph Scheit)
Date: Mon, 17 Dec 2007 18:12:11 +0100
Subject: [SciPy-user] read_array problem
In-Reply-To: <mailman.11.1197828008.20623.scipy-user@scipy.org>
References: <mailman.11.1197828008.20623.scipy-user@scipy.org>
Message-ID: <200712171812.11940.cscheit@lstm.uni-erlangen.de>

Hello everybody,

just for curiosity I have a question regarding read_array.
When I use the scipy.io read_array function I observe
some behaviour which I don't understand...

First I call the function like this:

self.varVector = read_array(f, columns=cols, lines=((0, 0 + bsize), ))

with cols = ((1,3), )
bsize = 5832
and I get an exception:
ValueError: Argument lines must be a sequence of integers and/or range tuples.

Calling now read_array like this

self.varVector = read_array(f, columns=cols, lines=((0, 0 + 5832), ))

works fine to my surprise...
cols is here the same like before.

Thanks in advance,

Christoph


From whenney at gmail.com  Mon Dec 17 14:01:34 2007
From: whenney at gmail.com (Will Henney)
Date: Mon, 17 Dec 2007 19:01:34 +0000 (UTC)
Subject: [SciPy-user] Mathematica Element-wise Multiplication
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>	<4765C29E.6070306@slac.stanford.edu>	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>	<476604B7.4060400@slac.stanford.edu>
	<e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
	<47662A12.7060107@slac.stanford.edu>
Message-ID: <loom.20071217T185744-481@post.gmane.org>

Johann Cohen-Tanugi <cohen <at> slac.stanford.edu> writes:
> hmmm...... okay... What I wanted was to transpose a 1D array into a 
> vector, or vice-versa, with the linear algebra behavior in mind. I 
> understand that numpy does not follow this, but I cannot believe that 
> this behavior *is* what everybody wants! Tom's initial email was 
> symptomatic, and Stefan's response, with the proposal to use the T 
> method even more so!

I think you want to be using numpy.matrix, rather than numpy.array: 

In [17]: m = numpy.matrix([1, 2, 3])

In [18]: m.T

Out[18]: 
matrix([[1],
        [2],
        [3]])

Cheers

Will






From stefan at sun.ac.za  Mon Dec 17 14:04:56 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Mon, 17 Dec 2007 21:04:56 +0200
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <47662A12.7060107@slac.stanford.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>
	<20071215113142.GH7167@mentat.za.net>
	<4765C29E.6070306@slac.stanford.edu>
	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>
	<476604B7.4060400@slac.stanford.edu>
	<e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
	<47662A12.7060107@slac.stanford.edu>
Message-ID: <20071217190455.GM21448@mentat.za.net>

On Sun, Dec 16, 2007 at 11:49:38PM -0800, Johann Cohen-Tanugi wrote:
> Matthieu Brucher wrote:
> >
> >
> > 2007/12/17, Johann Cohen-Tanugi <cohen at slac.stanford.edu 
> > <mailto:cohen at slac.stanford.edu>>:
> >
> >     thanks for these precisions, David. Reading it, I still come to think
> >     that it is a potential source of confusion to let a "row array" have a
> >     transpose or T method, that essentially does nothing.
> >
> >
> >
> > In object oriented code, this can happen often, but it is not a 
> > problem. It does what you want : inverse the axis, even if there is 
> > only one axis.
> hmmm...... okay... What I wanted was to transpose a 1D array into a 
> vector, or vice-versa, with the linear algebra behavior in mind. I 
> understand that numpy does not follow this, but I cannot believe that 
> this behavior *is* what everybody wants! Tom's initial email was 
> symptomatic, and Stefan's response, with the proposal to use the T 
> method even more so!

I'm not sure what my response was symptomatic of, but if you don't
like the behaviour of ndarrays, you may consider using 'matlib':

In [1]: from numpy import matlib

In [2]: x = matlib.matrix([1,2,3])

In [3]: x
Out[3]:
matrix([[1, 2, 3]])

In [4]: x.T
Out[4]:
matrix([[1],
        [2],
        [3]])

In [5]: x * x.T
Out[5]:
matrix([[14]])

> Assuming that this natural linear algebra could be retrieved when, and 
> *only* when, the array is 1D, I do not see how such an implementation 
> could break codes that depend on it, because I don't see why someone 
> would call 'a.T' just to have 'a' again.... But it is probably my lack 
> of imagination.

When I call "transpose", I don't expect an array to change dimensions.
Like Matthieu wrote, it swaps the axes, and if you have only one axis
there is nothing to swap.  Start with a two-dimensional array, and the
transpose will do what you expect.

Regards
St?fan


From lev at columbia.edu  Mon Dec 17 19:38:18 2007
From: lev at columbia.edu (Lev Givon)
Date: Mon, 17 Dec 2007 19:38:18 -0500
Subject: [SciPy-user] illegal instruction error in scipy.linalg.decomp.svd
Message-ID: <20071218003816.GE15380@localhost.cc.columbia.edu>

On a Pentium 4 Linux box running python 2.5.1, scipy 0.6.0, numpy
1.0.4, and lapack 3.0, I recently noticed that
scipy.linalg.decomp.svd() fails (and causes python to crash) with an
"illegal instruction" error. A bit of debugging revealed that the
error occurs in the line

lwork = calc_lwork.gesdd(gesdd.prefix,m,n,compute_uv)[1]

in scipy/linalg/decomp.py. Interestingly, scipy 0.5.2.1 on the same
box (with all of the other software packages unchanged) does not
exhibit this problem. Moreover, when I install scipy 0.6.0 along with
all of the other packages on a Linux machine containing a Pentium D
CPU, I do not observe the problem either.

Being that I am running Mandriva Linux, closed scipy bug 540 caught my
eye. I'm not sure how it could be related to the above problem, though
(and I also do not know what the lapack patch mentioned in the ticket
could have been - even though I have been maintaining Mandriva's
lapack lately :-).

Thoughts?

     	     	   	  	       	      	     L.G.


From david at ar.media.kyoto-u.ac.jp  Mon Dec 17 22:46:39 2007
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 18 Dec 2007 12:46:39 +0900
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <47662A12.7060107@slac.stanford.edu>
References: <b02c60870712142126h291902f8u1824db50191b4c8f@mail.gmail.com>	<20071215113142.GH7167@mentat.za.net>	<4765C29E.6070306@slac.stanford.edu>	<4765DEE0.9030401@ar.media.kyoto-u.ac.jp>	<476604B7.4060400@slac.stanford.edu>	<e76aa17f0712162248v14e06c1dx766eade384a6f9b8@mail.gmail.com>
	<47662A12.7060107@slac.stanford.edu>
Message-ID: <4767429F.5090802@ar.media.kyoto-u.ac.jp>

Johann Cohen-Tanugi wrote:
> Matthieu Brucher wrote:
>   
>> 2007/12/17, Johann Cohen-Tanugi <cohen at slac.stanford.edu 
>> <mailto:cohen at slac.stanford.edu>>:
>>
>>     thanks for these precisions, David. Reading it, I still come to think
>>     that it is a potential source of confusion to let a "row array" have a
>>     transpose or T method, that essentially does nothing.
>>
>>
>>
>> In object oriented code, this can happen often, but it is not a 
>> problem. It does what you want : inverse the axis, even if there is 
>> only one axis.
>>     
> hmmm...... okay... What I wanted was to transpose a 1D array into a 
> vector, or vice-versa, with the linear algebra behavior in mind. I 
> understand that numpy does not follow this, but I cannot believe that 
> this behavior *is* what everybody wants! Tom's initial email was 
> symptomatic, and Stefan's response, with the proposal to use the T 
> method even more so!
>
> Assuming that this natural linear algebra could be retrieved when, and 
> *only* when, the array is 1D, I do not see how such an implementation 
> could break codes that depend on it, because I don't see why someone 
> would call 'a.T' just to have 'a' again.... But it is probably my lack 
> of imagination.
>   
Indeed, it is a lack of imagination :) I can assure you that it would 
break a lot of my code, for example. I really understand your problem: 
you can see that on numpy ML that 18 months ago, I was exactly in the 
same situation as you. I used matlab a lot, was quite proficient in it, 
and was confused by this. But it really makes sense once you forget 
about matlab. I personally do not use numpy.matrix; not that it does not 
make sense to use them (when you really want to do linear algebra), but 
I think it just prevents you from using numpy at 100 %.
> Anyway, enough of this. I am sure the developers know better than me...
>   
IMHO, the 'problem' is coming from the fact that arrays are used for two 
things in matlab/numpy/etc....: for linear algebra, of course, but also, 
perhaps even more important, for speed reasons. For various reasons, it 
is really hard to do a fast script language. In particular when looping, 
and vectorizing code is a good way to avoid most of them. When for 
example I use a rank 2 array to compute some statistics on it, I do not 
really think about it as an array (the maths on it is not, for sure), 
but as a list of arrays; I use arrays because it is much faster.

If I had a choice, I would write the relevant code with loops (most of 
the time, it is clearer using loops). This is in clear contrast with 
linear algebra code where handling directly matrices and vectors is much 
clearer than doing loops manually. Matlab, for historical reasons I 
guess, is more geared towards linear algebra, whereas numpy is more 
geared toward 'using arrays for speed reasons'; in particular, 
broadcasting is much more powerful in numpy that it is in matlab.

cheers,

David


From wnbell at gmail.com  Mon Dec 17 23:41:32 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Tue, 18 Dec 2007 04:41:32 +0000 (UTC)
Subject: [SciPy-user] Sparse with fast element-wise multiply?
References: <27261D40-C6F1-402B-AFE3-D6B333CDA4FB@cs.toronto.edu>
	<80c99e790712161103j4f5ae3ecj4b53d6100403dea8@mail.gmail.com>
	<65C83323-C9E0-487A-8CFE-A5238B0A4C57@cs.toronto.edu>
Message-ID: <loom.20071218T042614-323@post.gmane.org>

David Warde-Farley <dwf <at> cs.toronto.edu> writes:

> Thanks for your reply. Actually, though, I'm not looking for matrix- 
> multiply; I know CSR/CSC is faster for that.
> 
> I'm looking for elementwise multiply, i.e. take two matrices of the  
> same size, multiply each element in matrix 1 with the corresponding  
> element in matrix 2, and put the result in the same position in the  
> result matrix (as with matlab's a .* b syntax). It seems lil_matrix is  
> the only sparse type that implements .multiply(), and converting to  
> lil_matrix is slow, so I've written my own function for multiplying  
> two coo_matrix's together. It's a LITTLE slower than .multiply() but  
> faster than converting both matrices to lil_matrix and then multiplying.
> 
> I'm not entirely sure how CSR and CSC work so it might be possible to  
> implement faster elementwise multiplication on top of them.
> 
> David
> 

Currently elementwise multiplication is exposed through A**B where A and B are
csr_matrix or csc_matrix objects.  You can expect similar performance to A+B.

I don't know why ** was chosen, it was that way before I started working on
scipy.sparse.  I've added a .multiply() method to the sparse matrix base class
that goes through csr_matrix:
http://projects.scipy.org/scipy/scipy/changeset/3682




From wnbell at gmail.com  Mon Dec 17 23:52:37 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Tue, 18 Dec 2007 04:52:37 +0000 (UTC)
Subject: [SciPy-user] Creating a sparse matrix
References: <web-93277839@uni-stuttgart.de> <47612D92.4020408@ntc.zcu.cz>
	<web-93278688@uni-stuttgart.de> <47613329.4020603@ntc.zcu.cz>
	<web-93282502@uni-stuttgart.de> <476153A7.1020707@ntc.zcu.cz>
	<web-93290641@uni-stuttgart.de>
Message-ID: <loom.20071218T044821-786@post.gmane.org>

Nils Wagner <nwagner <at> iam.uni-stuttgart.de> writes:

> Many thanks.
> The file contains only the upper triangular part of the
> symmetric matrix. How can I add the lower part to A ?

starting with row,col,data do the following:

mask = row != col #off-diagonal mask
new_row = concatenate((row,col[mask]))
new_col = concatenate((col,row[mask]))
new_data = concatenate((data,data[mask]))

This adds the transpose of the off-diagonal part of the matrix to the matrix.




From wnbell at gmail.com  Tue Dec 18 00:02:08 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Tue, 18 Dec 2007 05:02:08 +0000 (UTC)
Subject: [SciPy-user] scipy 0.6.0 ImportError: cannot import name cscmux
References: <338357900711180250u1fd08b86ide9a504d664e0cc1@mail.gmail.com>
Message-ID: <loom.20071218T050006-223@post.gmane.org>

youngsu park <youngsu999 <at> gmail.com> writes:
 
> 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]:

I believe sparsetools.pyd is from a previous scipy installation.  Deleting it
should work (as you've done).  To be safe I would delete /site-packages/scipy
and reinstall scipy.



From wnbell at gmail.com  Tue Dec 18 00:06:58 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Tue, 18 Dec 2007 05:06:58 +0000 (UTC)
Subject: [SciPy-user] scipy.sparse problem
References: <ffgpe1$jr7$1@ger.gmane.org>
Message-ID: <loom.20071218T050525-104@post.gmane.org>

Donald Fredkin <drfredkin <at> ucsd.edu> writes:

> Can anyone tell me how to fix this? It was OK with an earlier version
> of scipy.
> 
> This is, by the way, on Windows XP.

Try the steps at the bottom of this thread:
http://projects.scipy.org/scipy/scipy/ticket/516





From wnbell at gmail.com  Tue Dec 18 00:16:08 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Tue, 18 Dec 2007 05:16:08 +0000 (UTC)
Subject: [SciPy-user] sparse matrix dtype
References: <b14b627d0710090636u1267394bp11f11e9171f61be0@mail.gmail.com>
	<b14b627d0710090703q54089ad5od9c79af95b30c789@mail.gmail.com>
Message-ID: <loom.20071218T051301-639@post.gmane.org>

Robin <robince <at> gmail.com> writes:

> 
> On 10/9/07, Robin <robince <at> gmail.com> wrote:
> I am trying to make a large sparse matrix - the values are all integers (in
fact all non-zeros will be 1) so it would save me a lot of memory if I could use
dtype=byte.
> 
> I added 'b' to the string of allowed dtypes in  getdtype() on line  2791 of 
> sparse.py.It now seems to behave as I would expect (hope), but it can't be
that simple can it?Is it likely that doing this will break something else? Why
are the dtypes restricted in the first place.ThanksRobin
> 

FWIW integer dtypes are supported in the next release (and the current SVN)
http://projects.scipy.org/scipy/scipy/ticket/225

The problem before was not on the python side.  It was the backend code
(sparse.sparsetools) that didn't support integer types.



From amg at iri.columbia.edu  Tue Dec 18 10:51:24 2007
From: amg at iri.columbia.edu (Arthur M. Greene)
Date: Tue, 18 Dec 2007 10:51:24 -0500
Subject: [SciPy-user] Estimating AR(1) model?
In-Reply-To: <b11ea23c0712081619h638fba46sdc4ebcb6e15d803@mail.gmail.com>
References: <b11ea23c0712081619h638fba46sdc4ebcb6e15d803@mail.gmail.com>
Message-ID: <4767EC7C.8090600@iri.columbia.edu>

Well, it's not scipy, but this small Python package has a convenient 
routine for estimating the parameters of an autoregressive process:

http://geosci.uchicago.edu/~cdieterich/span/

It is based on Numeric; you'll have to probe a little to figure out what 
algorithm is used for the estimation.

Cheers,

AMG

Webb Sprague wrote:
> Hi all,
> 
> Is there a quick approach for estimating a simple AR(1) time series
> model from a vector of data points?  Currently I use lingregress on
> lagged data and roll my own standard error computations.  It would be
> great if someone smarter than I has done this already, especially if
> they use MLE or Yule-Walker.
> 
> I am using Gentoo, scipy.__version__ == '0.6.0', and I can't find a
> scipy/stats/models directory in my installation, if that matters.
> 
> Thx
> W
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user





From timmichelsen at gmx-topmail.de  Tue Dec 18 15:42:25 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Tue, 18 Dec 2007 21:42:25 +0100
Subject: [SciPy-user] flagging no data in timeseries
Message-ID: <fk9bbk$5ri$1@ger.gmane.org>

Hello,
I while ago I was pointed to the timeseries module from sandbox by one 
of its developers. Thanks again for that. I am still having some 
troubles and questions:

Since I am still very new to the package I need some hints.
In my original data nodata values are marked with "-999".
How can I import the data or create the time series and exclude these no 
data points from further processing?

I presume it has to do with the mask keyword. But I cannot get it.

Here is the code:

data = N.loadtxt(mydata_file, comments='#', delimiter='\t', 
converters=None, skiprows=2, usecols=None, unpack=False)
# define start date
D_hr = TS.Date(freq='HR', year=1995, month=2, day=1, hour=1)
# mask
## prepare time series
# take only needed values
myvalues = data[:,5]
# hourly values
myvalues_ts_hourly = TS.time_series(myvalues, start_date=D_hr)

I will appreciate every help.

For further readability and archivability I write my other questions in 
separate emails.

Kind regards,
Timmie



From timmichelsen at gmx-topmail.de  Tue Dec 18 16:09:05 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Tue, 18 Dec 2007 22:09:05 +0100
Subject: [SciPy-user] get result array after timeseries operations
Message-ID: <fk9ctk$bg1$1@ger.gmane.org>

Hi,
this is meant as a continuation to

How can one save the read the result of time series operations into a array?

For instance, if I convert data in an hourly frequency to daily averages 
how to I read the daily averages into a array for further processing?

when I print out my daily timeseries converted from hourly data I get 
something like this:

In: myvalues_ts_daily
Out:
timeseries([  1.4   89.4  3.5 ...,  11.5  1.6
     0.        ],
            dates = [01-Feb-1995 ... 01-Dec-2006],
            freq  = D)

What I would like is an array with just the values of the daily averages .
Additional a report-like array output with the format
day value
1   3
2   11

would be nice.

How can I achieve this?

Do I need to use the report functions [1]?

Thanks again for the ice package. Things are looking nice!

Regards,
Timmie

[1] Reports - 
http://www.scipy.org/SciPyPackages/TimeSeries#head-e7ce84a06d526536a962deb356ba6258e52b391d 



From whenney at gmail.com  Wed Dec 19 01:10:59 2007
From: whenney at gmail.com (William Henney)
Date: Wed, 19 Dec 2007 06:10:59 +0000 (UTC)
Subject: [SciPy-user] Mayavi2 installation woes (OS X 10.4.11, Python 2.5)
Message-ID: <loom.20071219T053615-624@post.gmane.org>

Hi list,

I have been trying to install MayaVi2 on an intel Mac running 10.4.11, but I am
running into a brick wall. 

I have been trying to follow the instructions I have found in various threads on
this list and at https://svn.enthought.com/enthought/wiki/IntelMacPython25. I
have managed to install all the dependencies (vtk, pyrex, wxpython, etc) but I
just can't get the enthought.traits package to build. 

I have tried the versions from the SVN repo and also the tar balls from
http://code.enthought.com/downloads/source/ets2.6/ but I always get the error
when building enthought.traits:

TypeError: swig_sources() takes exactly 3 arguments (2 given)

An example of a full stack trace is given below. It seems that the problem has
something to do with setuptools, and possibly with pyrex. I was initially using
setuptools 0.6c7 and Pyrex 0.9.6.4, but after reading negative comments about
0.9.6, I downgraded to 0.9.5.1a, although this makes no difference. 

I have also tried using the latest development version of setuptools
(0.7a1dev-r58661), but this doesn't work at all due to problems with "Namespace
Packages" (with 0.6 these just give a warning, as seen in the output below). 

Does anyone have any idea how to solve this?

Cheers

Will Henney 


$ python setup.py build
running build
running build_py
WARNING: enthought is a namespace package, but its __init__.py does
not declare_namespace(); setuptools 0.7 will REQUIRE this!
(See the setuptools manual under "Namespace Packages" for details.)

WARNING: enthought.traits is a namespace package, but its __init__.py does
not declare_namespace(); setuptools 0.7 will REQUIRE this!
(See the setuptools manual under "Namespace Packages" for details.)

WARNING: enthought.traits.ui is a namespace package, but its __init__.py does
not declare_namespace(); setuptools 0.7 will REQUIRE this!
(See the setuptools manual under "Namespace Packages" for details.)

running egg_info
writing requirements to enthought.traits.egg-info/requires.txt
writing enthought.traits.egg-info/PKG-INFO
writing namespace_packages to enthought.traits.egg-info/namespace_packages.txt
writing top-level names to enthought.traits.egg-info/top_level.txt
writing dependency_links to enthought.traits.egg-info/dependency_links.txt
reading manifest file 'enthought.traits.egg-info/SOURCES.txt'
writing manifest file 'enthought.traits.egg-info/SOURCES.txt'
running build_ext
building 'enthought.traits.ctraits' extension
Traceback (most recent call last):
  File "setup.py", line 72, in <module>
    zip_safe = False,
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/core.py",
line 151, in setup
    dist.run_commands()
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/dist.py",
line 974, in run_commands
    self.run_command(cmd)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/dist.py",
line 994, in run_command
    cmd_obj.run()
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/command/build.py",
line 112, in run
    self.run_command(cmd_name)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/cmd.py",
line 333, in run_command
    self.distribution.run_command(command)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/dist.py",
line 994, in run_command
    cmd_obj.run()
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
site-packages/setuptools-0.6c7-py2.5.egg/setuptools/command/build_ext.py",
line 46, in run
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/command/build_ext.py",
line 290, in run
    self.build_extensions()
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
site-packages/Pyrex/Distutils/build_ext.py",
line 82, in build_extensions
    self.build_extension(ext)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
site-packages/setuptools-0.6c7-py2.5.egg/setuptools/command/build_ext.py",
line 175, in build_extension
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
distutils/command/build_ext.py",
line 453, in build_extension
    sources = self.swig_sources(sources, ext)
  File
"/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/
site-packages/setuptools-0.6c7-py2.5.egg/setuptools/command/build_ext.py",
line 77, in swig_sources
TypeError: swig_sources() takes exactly 3 arguments (2 given)




From fie.pye at atlas.cz  Wed Dec 19 13:18:01 2007
From: fie.pye at atlas.cz (Pye Fie)
Date: Wed, 19 Dec 2007 18:18:01 GMT
Subject: [SciPy-user] FW: Re:  Building VTK
Message-ID: <90ec234f72464a65b7cf0ae9a1e9398e@837e8f6af700492f9509f88823a3a0c0>

Hi.

Finaly I managed to build and install VTK.

Solution was to call g++34 insted of gcc34 
CMAKE_CXX_COMPILER:FILEPATH=/usr/bin/g++34

fie pye


>---------------------------------------------------------
>Od: Pye Fie
>P?ijato: 12.12.2007 23:59:38
>P?edm?t: Re: [SciPy-user] Building VTK
>
> 
>Hi Metthieu.
>Than you for your response. I posted my question on VTK ML first.
>http://public.kitware.com/pipermail/vtkusers/2007-December/093627.html
>Unfortunately I didn't get any answer until now. 
>
>You asker about my compiler. 
>CentOS 5.0 is compiled with gcc 4.1. I installed also gcc 4.2 and gcc 3.4.6 on the PC.
>
>I would like to have a computational environment based on Python and its module. I compiled Python and module Ipython, SciPy, NumPy, tables, matplotlib, wxpython ......... and libraries such as BLAS, LAPACK, ATLAS, HDF5, ........ with gcc3.4.6. Now I continue with compiling VTK with gcc3.4.6 too. More information about compilation setup is the file CMakeCache.txt.
>
>Best regards.
>Fie pye
>
>
>Od: Matthieu Brucher
>P?ijato: 11.12.2007 21:18:04
>P?edm?t: Re: [SciPy-user] Building VTK
>Besides, you tell us you use GCC 3.4.6 but in your description, it is GCC 4.2...
>The issue may be solved by linking against GCC libraries, but again, the VTK ML may be a better place for the solution.
>Matthieu
>2007/12/11, Matthieu Brucher <matthieu.brucher at gmail.com>: Hi,
>This issue is related to a C++ problem in VTK, not Python, so we can't say much. I advise you to post this on the VTK ML.
>Matthieu
>2007/12/11, Pye Fie < fie.pye at atlas.cz>:Hello.
>I would like to build and install VTK and MayaVi2. 
>PC: 2x Dual-core Opteron
>275
>OS: CentOS 5.0, kernel 2.6.18-8.1.15.el5
>OS compiler: gcc version 4.2.0, gfortran
>I
>have already built and installed python2.5.1 and other module with compiler gcc3.4.6 . 
>Now I can't build VTK. I am building with gcc3.4.6. Building stops on the following
>error:
>[ 74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkZLibDataCompressorTcl.o
>[
>74%] Building CXX object IO/CMakeFiles/vtkIOTCL.dir/vtkIOTCLInit.o 
>Linking CXX shared
>library ../bin/libvtkIOTCL.so
>[ 74%] Built target vtkIOTCL
>Scanning dependencies
>of target vtkParseOGLExt
>[ 74%] Building CXX object Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/ParseOGLExt.o 
>[
>74%] Building CXX object Utilities/ParseOGLExt/CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o
>Linking
>CXX executable ../../bin/vtkParseOGLExt
>CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o:
>In function `std::__verify_grouping(char const*, unsigned long, std::basic_string
>std::char_traits, std::allocator > const&)':
>Tokenizer.cxx:(.text+0x19):
>undefined reference to `std::basic_string, std::allocator 
>>::size() const'
>Tokenizer.cxx:(.text+0x70): undefined reference to `std::basic_string
>std::char_traits, std::allocator >::operator[](unsigned long) const'
>Tokenizer.cxx : (.text+0xb0):
>undefined reference to `std::basic_string, std::allocator
>>::operator[](unsigned long) const'
>Tokenizer.cxx:(.text+0xdd): undefined reference 
>to `std::basic_string, std::allocator >::operator[](unsigned
>long) const'
>CMakeFiles/vtkParseOGLExt.dir/Tokenizer.o: In function `Tokenizer::Tokenizer(char 
>const*, char const*)':
>Tokenizer.cxx:(.text+0x12a): undefined reference to `std::allocator::allocator()'
>Tokenizer.cxx:(.text+0x13b):
>undefined reference to `std::basic_string, std::allocator 
>>::basic_string(char const*, std::allocator const&)'
>Tokenizer.cxx:(.text+0x14e):
>undefined reference to `std::allocator::~allocator()'
>Tokenizer.cxx:(.text+0x164):
>undefined reference to `std::allocator::~allocator()' 
>Tokenizer.cxx:(.text+0x16d):
>undefined reference to `std::allocator::allocator()'
>Tokenizer.cxx:(.text+0x182):
>undefined reference to `std::basic_string, std::allocator 
>>::basic_string(char const*, std::allocator const&)'
>The list continues. It
>seams that there is something wrong with OpenGL but I can't find what is wrong. Could
>anybody help me? 
>I have attached more detailed information about VTK configuration,
>building output and building script.
>Best regards
>fie pye
>------------------------------------------
>http://search.atlas.cz/
>_______________________________________________
>SciPy-user mailing list
>SciPy-user at scipy.org
>http://projects.scipy.org/mailman/listinfo/scipy-user
>-- 
>French PhD student
>Website :  http://matthieu-brucher.developpez.com/ 
>Blogs : http://matt.eifelle.com and  http://blog.developpez.com/?blog=92
>LinkedIn :  http://www.linkedin.com/in/matthieubrucher 
>-- 
>French PhD student
>Website : http://matthieu-brucher.developpez.com/
>Blogs :  http://matt.eifelle.com and http://blog.developpez.com/?blog=92
>LinkedIn : http://www.linkedin.com/in/matthieubrucher  
>
>------------------------------------------
>
>
>
>http://search.atlas.cz/

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

http://search.atlas.cz/
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: attachment2.txt
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071219/73c6faa6/attachment.txt>

From dwf at cs.toronto.edu  Wed Dec 19 16:21:08 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Wed, 19 Dec 2007 16:21:08 -0500
Subject: [SciPy-user] arraysetops & string arrays
Message-ID: <8C1A7051-36CA-4B8E-A1F0-0E4C6834BD7C@cs.toronto.edu>

Hi there,

I noticed that setdiff1d() fails on string arrays.  (numpy 1.0.3.1).  
There's nothing in the documentation that suggests it's for numeric  
dtypes only; should there be, or is this a bug?

The problem appears to be the line

tt = nm.concatenate( (zlike( ar1 ), zlike( ar2 ) + 1) )

where zlike() is returning an array of empty strings, and thus adding  
1 is causing problems. I tried modifying it to cast to int32 without  
much success.

David


From erik.volz at gmail.com  Wed Dec 19 19:32:12 2007
From: erik.volz at gmail.com (Erik Volz)
Date: Wed, 19 Dec 2007 16:32:12 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
Message-ID: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>

There are several aspects of scipy that are counterintuitive to typical
users. Tranpose of 1D arrays is a big one. Another example is
z+=z.T # where dimension z > 1
which does not do what a naive user would think it does (try it).

I recently introduced very damaging bugs into my code by making both of
these mistakes; it took a long time for me to track down the problem. Of
course, the behavior is very logical from a developer's perspective, but the
veterans of this list have been very unsympathetic. At a minimum, there
should be warnings in the doc-strings about this behavior.

Erik Volz


From: Johann Cohen-Tanugi <cohen <at> slac.stanford.edu>

> Subject: Re: Mathematica Element-wise Multiplication<http://news.gmane.org/find-root.php?message_id=%3c47662A12.7060107%40slac.stanford.edu%3e>
> Newsgroups: gmane.comp.python.scientific.user<http://news.gmane.org/gmane.comp.python.scientific.user>
> Date: 2007-12-17 07:49:38 GMT (2 days, 16 hours and 31 minutes ago)
>
> Matthieu Brucher wrote:
> >
> >
> > 2007/12/17, Johann Cohen-Tanugi <cohen <at> slac.stanford.edu
> > <mailto:cohen <at> slac.stanford.edu>>:
> >
> >     thanks for these precisions, David. Reading it, I still come to think
> >     that it is a potential source of confusion to let a "row array" have a
> >     transpose or T method, that essentially does nothing.
> >
> >
> >
> > In object oriented code, this can happen often, but it is not a
> > problem. It does what you want : inverse the axis, even if there is
> > only one axis.
> hmmm...... okay... What I wanted was to transpose a 1D array into a
> vector, or vice-versa, with the linear algebra behavior in mind. I
> understand that numpy does not follow this, but I cannot believe that
> this behavior **is** what everybody wants! Tom's initial email was
> symptomatic, and Stefan's response, with the proposal to use the T
> method even more so!
>
> Assuming that this natural linear algebra could be retrieved when, and
> **only** when, the array is 1D, I do not see how such an implementation
> could break codes that depend on it, because I don't see why someone
> would call 'a.T' just to have 'a' again.... But it is probably my lack
> of imagination.
> Anyway, enough of this. I am sure the developers know better than me...
>
> cheers,
> Johann
>
> >
> >
> >     I guess there is a
> >     reason behind this situation, but given the fact that it is there,
> >     I am
> >     wondering whether T or transpose of a row array could in fact return
> >     what it is expected to, aka the 2d column array. Is there any
> >     reason not
> >     to have this functionality?
> >
> >
> >
> > More code in a simple function (thus slower) ? Breaking some code that
> > depend on it ? Not doing what the documentation says ?
> > I think the method should not be changed, it does exactly what it says
> > it does. Changing its behaviour because of Matlab is not a good
> > solution, Python is not Matlab.
> >
> > Matthieu
> > --
> > French PhD student
> > Website :
> http://matthieu-brucher.developpez.com/
> > Blogs : http://matt.eifelle.com <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
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071219/07e035bb/attachment.html>

From oliphant at enthought.com  Wed Dec 19 20:26:35 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Wed, 19 Dec 2007 19:26:35 -0600
Subject: [SciPy-user] Mailing list was not sending out new mails for awhile.
Message-ID: <4769C4CB.7080609@enthought.com>


Hi all,

The postfix service on the server hosting several of the mailing lists 
was down since Monday.  Mails to the list were preserved and archived 
but were not being distributed to subscribers.   We restarted the 
postfix service and messages should now be going out.

Apologies for the problem.

-Travis O.




From oliphant at enthought.com  Wed Dec 19 20:50:20 2007
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Wed, 19 Dec 2007 19:50:20 -0600
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
References: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
Message-ID: <4769CA5C.70205@enthought.com>

Erik Volz wrote:
> There are several aspects of scipy that are counterintuitive to 
> typical users. Tranpose of 1D arrays is a big one. Another example is
> z+=z.T # where dimension z > 1
> which does not do what a naive user would think it does (try it).
>
> I recently introduced very damaging bugs into my code by making both 
> of these mistakes; it took a long time for me to track down the 
> problem. Of course, the behavior is very logical from a developer's 
> perspective, but the veterans of this list have been very 
> unsympathetic. At a minimum, there should be warnings in the 
> doc-strings about this behavior.

Yes, absolutely, warnings should be there in the docstring.  Your 
perspective is not illogical nor are many of us unsympathetic to it.   
But,  following your rule would be an "exceptional" case  to an 
otherwise easy to explain rule. 

Our approach thus far has been to educate users coming from other 
languages that there is such a thing as a 1D array that is neither a 
"row" nor a "column" vector and so in fact a transpose operation is not 
really what you meant (as it does nothing to a 1D array).   Instead what 
you wanted to do was convert the 1-D array to a 2-D array with 1 row and 
then transpose that to a 2D array with 1 column.  

Doing this "automatically" for you is assuming a particular convention 
which we at least try not to do that  often.  The problems it causes, 
come down to introducing oddities in code that could otherwise work over 
arbitrary dimensions.

Finding the right balance between least-surprise and rule-exception 
headaches is not an easy one (nor will I ever claim that NumPy has made 
all the right choices).  Because NumPy is community developed, there is 
always the possibility that the behavior will change.  Of course, after 
something has come into use, it needs more of an argument than 
subjective points about a typical user (although I like hearing those 
kinds of suggestions when deciding what a *new* or *revamped* thing 
should do).  

Thanks for sharing your perspective.  I'm very aware that for every 
piece of "negative" feedback (and yours really wasn't) there are 100 
people who feel the same way but don't have the time to tell us.   I'm 
very sorry to hear of the bugs you couldn't find because of this 
mis-communication in the .T attribute.

Best regards,

-Travis Oliphant




From robert.kern at gmail.com  Wed Dec 19 21:01:33 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 19 Dec 2007 20:01:33 -0600
Subject: [SciPy-user] Mayavi2 installation woes (OS X 10.4.11,
 Python 2.5)
In-Reply-To: <loom.20071219T053615-624@post.gmane.org>
References: <loom.20071219T053615-624@post.gmane.org>
Message-ID: <4769CCFD.4080301@gmail.com>

William Henney wrote:
> Hi list,
> 
> I have been trying to install MayaVi2 on an intel Mac running 10.4.11, but I am
> running into a brick wall. 

Can you join us over on enthought-dev? It's a more appropriate forum.

  https://mail.enthought.com/mailman/listinfo/enthought-dev

> I have been trying to follow the instructions I have found in various threads on
> this list and at https://svn.enthought.com/enthought/wiki/IntelMacPython25. I
> have managed to install all the dependencies (vtk, pyrex, wxpython, etc) but I
> just can't get the enthought.traits package to build. 
> 
> I have tried the versions from the SVN repo and also the tar balls from
> http://code.enthought.com/downloads/source/ets2.6/ but I always get the error
> when building enthought.traits:
>
> TypeError: swig_sources() takes exactly 3 arguments (2 given)
> 
> An example of a full stack trace is given below. It seems that the problem has
> something to do with setuptools, and possibly with pyrex. I was initially using
> setuptools 0.6c7 and Pyrex 0.9.6.4, but after reading negative comments about
> 0.9.6, I downgraded to 0.9.5.1a, although this makes no difference. 

Can you double-check that the Pyrex that you can import is the version that you
think it is? One of the changes in 0.9.6 causes exactly this problem.

>>> from Pyrex.Compiler import Version
>>> Version.version
'0.9.5.1'

> I have also tried using the latest development version of setuptools
> (0.7a1dev-r58661), but this doesn't work at all due to problems with "Namespace
> Packages" (with 0.6 these just give a warning, as seen in the output below). 

The tarballs you grabbed are experimental ones for people trying to package ETS
for Linux distributions. The people making them wanted to remove the namespace
package features, but it appears that they did so incompletely.

The tarballs that you want are here:

  http://code.enthought.com/enstaller/eggs/source/

Alternately, we have a script that can check out the SVN for a given ETS package
and all of its dependencies.

  https://svn.enthought.com/svn/enthought/branches/ets_scripts_0.1

Install that like normal ("python setup.py install"). Then you will have a
script called "etsco":

$ cd ~/src
$ etsco enthought.mayavi
Initializing list of projects found in svn repositories.
    Processing repository: https://svn.enthought.com/svn/enthought
Finished initializing svn repository projects.
    Checking out "enthought.util" source from
"https://svn.enthought.com/svn/enthought/tags/enthought.util_2.0.1b2"
A    /Users/rkern/src/enthought.mayavi_2.0.3a1/enthought.util_2.0.1b2/LICENSE.txt
A    /Users/rkern/src/enthought.mayavi_2.0.3a1/enthought.util_2.0.1b2/extras.map
....
A
/Users/rkern/src/enthought.mayavi_2.0.3a1/enthought.persistence_2.1.0a1/setup.cfg
 U   /Users/rkern/src/enthought.mayavi_2.0.3a1/enthought.persistence_2.1.0a1
Checked out revision 16353.

All project source has been checked out to
"/Users/rkern/src/enthought.mayavi_2.0.3a1"

WARNING: YOUR CHECKOUT INCLUDES PROJECTS FROM SVN TAGS -- PLEASE BE CAREFUL WHEN
CHECKING IN CHANGES AS TAGS SHOULD NOT BE CHANGED!  You can svn copy a tag to a
 branch, then svn switch to that, and then check in your changes for a future
release.

Now, in enthought.mayavi_2.0.3a1/ you have all of the current SVN for
enthought.mayavi and everything from ETS that it needs.

-- 
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 dwf at cs.toronto.edu  Thu Dec 20 03:56:27 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 20 Dec 2007 03:56:27 -0500
Subject: [SciPy-user] Sparse with fast element-wise multiply?
In-Reply-To: <loom.20071218T042614-323@post.gmane.org>
References: <27261D40-C6F1-402B-AFE3-D6B333CDA4FB@cs.toronto.edu>
	<80c99e790712161103j4f5ae3ecj4b53d6100403dea8@mail.gmail.com>
	<65C83323-C9E0-487A-8CFE-A5238B0A4C57@cs.toronto.edu>
	<loom.20071218T042614-323@post.gmane.org>
Message-ID: <B9361C61-2D5E-455D-A567-8E091C5832A9@cs.toronto.edu>

On 17-Dec-07, at 11:41 PM, Nathan Bell wrote:

> Currently elementwise multiplication is exposed through A**B where A  
> and B are
> csr_matrix or csc_matrix objects.  You can expect similar  
> performance to A+B.

Whoa, you're not kidding:

In [19]: time multiply_coo(k1,k2)
CPU times: user 0.77 s, sys: 0.08 s, total: 0.85 s
Wall time: 0.86
Out[19]:
<7083x7083 sparse matrix of type '<type 'numpy.float64'>'
	with 24226 stored elements in COOrdinate format>

In [20]: time k1csr ** k2csr
CPU times: user 0.02 s, sys: 0.00 s, total: 0.02 s
Wall time: 0.02
Out[20]:
<7083x7083 sparse matrix of type '<type 'numpy.float64'>'
	with 24226 stored elements in Compressed Sparse Row format>


Actually it's about 5 times FASTER than adding the two of them.  
Probably because in the latter it's the union of the elements that is  
the result, rather than the (typically sparser) intersection.

> I don't know why ** was chosen, it was that way before I started  
> working on
> scipy.sparse.

It seems sensible enough to me; I don't know how often I've had to  
exponentiate a matrix (much less a sparse one) and if you want to do  
any serious exponentiation it's usually cheaper (and I'd expect more  
numerically stable) to diagonalize it once & exponentiate the  
eigenvalues.
Is this ** behaviour documented anywhere?

>  I've added a .multiply() method to the sparse matrix base class
> that goes through csr_matrix:
> http://projects.scipy.org/scipy/scipy/changeset/3682


Muchos gracias.

DWF


From wnbell at gmail.com  Thu Dec 20 04:21:21 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Thu, 20 Dec 2007 03:21:21 -0600
Subject: [SciPy-user] Sparse with fast element-wise multiply?
In-Reply-To: <B9361C61-2D5E-455D-A567-8E091C5832A9@cs.toronto.edu>
References: <27261D40-C6F1-402B-AFE3-D6B333CDA4FB@cs.toronto.edu>
	<80c99e790712161103j4f5ae3ecj4b53d6100403dea8@mail.gmail.com>
	<65C83323-C9E0-487A-8CFE-A5238B0A4C57@cs.toronto.edu>
	<loom.20071218T042614-323@post.gmane.org>
	<B9361C61-2D5E-455D-A567-8E091C5832A9@cs.toronto.edu>
Message-ID: <d05265cb0712200121o1336c15ex3110b64e8a011302@mail.gmail.com>

On Dec 20, 2007 2:56 AM, David Warde-Farley <dwf at cs.toronto.edu> wrote:
> Is this ** behaviour documented anywhere?

I don't believe so.

In any case, I've made ** implement exponentiation as it does for
dense matrices.  You'll get a deprecation warning for (sparse **
sparse) in SciPy 0.7 telling you to use sparse.multiply(sparse)
instead.

-- 
Nathan Bell wnbell at gmail.com


From Alexander.Dietz at astro.cf.ac.uk  Thu Dec 20 04:58:01 2007
From: Alexander.Dietz at astro.cf.ac.uk (Alexander Dietz)
Date: Thu, 20 Dec 2007 09:58:01 +0000
Subject: [SciPy-user] Questions to optimize.leastsq
Message-ID: <9cf809a00712200158x1d7348b0hc5a79eeaec2a6878@mail.gmail.com>

Hi,

I am using optimize.leastsq for the first time, and I seem to have some
trouble with it. I am following the example given on the scipy webpage to
fit a either 3 or 4 parameter function. The fit itself works fine, and I get
reasonable results.
My problem is to extract the errors on my 3 or 4 fitting parameters. Here is
what I do exactly:

q,cov,w,e, success = optimize.leastsq(errfunc, p[:], args = (xData, yData),
full_output=1 )
if success==1:
    sigma = sqrt(diag(cov))


For the errors on my fitting parameters 'q' I take the square-root in the
diagonal elements of the covariance matrix. For one or two parameters I get
reasonable errors (i.e. the relative error is in the range of 30%), but for
other parameters the errors are a million times larger than the parameter
itself!

Example:
  bestfit parameter 0: -1.743351 +- 3226113.032304
  bestfit parameter 1: 0.778228 +- 0.145904
  bestfit parameter 2: 4.261635 +- 48185218.007348
  bestfit parameter 3: 21183.401658 +- 23799.387162

I checked the function I am fitting and when varying one of these parameters
everything looks 'ok' (i.e. no sudden change of the values in the order of
millions or so). I my sense, these errors (except parameter 1) do not make
sense.

So how can I fit a function to data and obtain reasonable errors? Maybe I
need to set one of the extra options in leastsq? I would appreciate any help
in this.


Cheers
  Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071220/1fa36d77/attachment.html>

From cimrman3 at ntc.zcu.cz  Thu Dec 20 05:31:41 2007
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Thu, 20 Dec 2007 11:31:41 +0100
Subject: [SciPy-user] arraysetops & string arrays
In-Reply-To: <8C1A7051-36CA-4B8E-A1F0-0E4C6834BD7C@cs.toronto.edu>
References: <8C1A7051-36CA-4B8E-A1F0-0E4C6834BD7C@cs.toronto.edu>
Message-ID: <476A448D.3040206@ntc.zcu.cz>

Hi David!

David Warde-Farley wrote:
> Hi there,
> 
> I noticed that setdiff1d() fails on string arrays.  (numpy 1.0.3.1).  
> There's nothing in the documentation that suggests it's for numeric  
> dtypes only; should there be, or is this a bug?
> 
> The problem appears to be the line
> 
> tt = nm.concatenate( (zlike( ar1 ), zlike( ar2 ) + 1) )
> 
> where zlike() is returning an array of empty strings, and thus adding  
> 1 is causing problems. I tried modifying it to cast to int32 without  
> much success.

Fixed in SVN. It was more a missing feature than a bug - I did not 
realized that it should work for non-numerical arrays, too.

If you do not use SVN, I can mail you the fix, just tell me.

If other functions have the same problems, let me know, please.

r.



From dmitrey.kroshko at scipy.org  Thu Dec 20 06:30:58 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Thu, 20 Dec 2007 13:30:58 +0200
Subject: [SciPy-user] Could anyone send me copy Python Magazine 2007/11?
Message-ID: <476A5272.2060100@scipy.org>

hi all,
could anyone send me copy of the journal Python Magazine 2007/11 (I mean 
zipped file) mentioned in scipy.org main page? Or provide an url where 
it could be downloaded.
There is only 2007/10 available for free download.

Thank you in advance, D.


From matthieu.brucher at gmail.com  Thu Dec 20 06:57:18 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 20 Dec 2007 12:57:18 +0100
Subject: [SciPy-user] Could anyone send me copy Python Magazine 2007/11?
In-Reply-To: <476A5272.2060100@scipy.org>
References: <476A5272.2060100@scipy.org>
Message-ID: <e76aa17f0712200357p7bb6b7e8k8d7e0ac3d14f9646@mail.gmail.com>

Hi,

The issue is password-protected and the mail of the subscriber is included
inside the pdf, so I don't think you will get one. Sorry...

Matthieu

2007/12/20, dmitrey <dmitrey.kroshko at scipy.org>:
>
> hi all,
> could anyone send me copy of the journal Python Magazine 2007/11 (I mean
> zipped file) mentioned in scipy.org main page? Or provide an url where
> it could be downloaded.
> There is only 2007/10 available for free download.
>
> Thank you in advance, D.
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://matthieu-brucher.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/20071220/aeac07c9/attachment.html>

From dmitrey.kroshko at scipy.org  Thu Dec 20 07:01:01 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Thu, 20 Dec 2007 14:01:01 +0200
Subject: [SciPy-user] Could anyone send me copy Python Magazine 2007/11?
In-Reply-To: <e76aa17f0712200357p7bb6b7e8k8d7e0ac3d14f9646@mail.gmail.com>
References: <476A5272.2060100@scipy.org>
	<e76aa17f0712200357p7bb6b7e8k8d7e0ac3d14f9646@mail.gmail.com>
Message-ID: <476A597D.3080709@scipy.org>

Hi Matthieu,
thank you for the tip, I'll wait till free version will be available.
Regards, D.

Matthieu Brucher wrote:
> Hi,
>
> The issue is password-protected and the mail of the subscriber is 
> included inside the pdf, so I don't think you will get one. Sorry...
>
> Matthieu
>
> 2007/12/20, dmitrey < dmitrey.kroshko at scipy.org 
> <mailto:dmitrey.kroshko at scipy.org>>:
>
>     hi all,
>     could anyone send me copy of the journal Python Magazine 2007/11
>     (I mean
>     zipped file) mentioned in scipy.org <http://scipy.org> main page?
>     Or provide an url where
>     it could be downloaded.
>     There is only 2007/10 available for free download.
>
>     Thank you in advance, D.
>     _______________________________________________
>     SciPy-user mailing list
>     SciPy-user at scipy.org <mailto:SciPy-user at scipy.org>
>     http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>
> -- 
> French PhD student
> Website : http://matthieu-brucher.developpez.com/ 
> <http://matthieu-brucher.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
>   



From pgmdevlist at gmail.com  Thu Dec 20 11:37:34 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Thu, 20 Dec 2007 11:37:34 -0500
Subject: [SciPy-user] flagging no data in timeseries
In-Reply-To: <fk9bbk$5ri$1@ger.gmane.org>
References: <fk9bbk$5ri$1@ger.gmane.org>
Message-ID: <200712201137.35658.pgmdevlist@gmail.com>

Hello,

> Since I am still very new to the package I need some hints.
> In my original data nodata values are marked with "-999".
> How can I import the data or create the time series and exclude these no
> data points from further processing?

The following solutions should be equivalent:

* use masked_where from maskedarray
>>>myvalues_ts_hourly = masked_where(myvalues_ts_hourly , -999)

* Use indexing
>>>myvalues_ts_hourly[myvalues_ts_hourly==-999] = M.masked



From pgmdevlist at gmail.com  Thu Dec 20 11:43:27 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Thu, 20 Dec 2007 11:43:27 -0500
Subject: [SciPy-user] get result array after timeseries operations
In-Reply-To: <fk9ctk$bg1$1@ger.gmane.org>
References: <fk9ctk$bg1$1@ger.gmane.org>
Message-ID: <200712201143.28092.pgmdevlist@gmail.com>

On Tuesday 18 December 2007 16:09:05 Tim Michelsen wrote:
> Hi,
> this is meant as a continuation to

> For instance, if I convert data in an hourly frequency to daily averages
> how to I read the daily averages into a array for further processing?

* possibility #1: use the keyword func while converting.
avgdata = convert(data,'D', func=maskedarray.mean)

* possibility #2:
If you don't use the keyword func, you end up with a 2d array, each row being 
a day, each column an hour. Just use maskedarray.mean on each row
avgdata = convert(data,'D').mean(-1)

If you only want the values, use the .series attribute, it will give you a 
view of the array as a MaskedArray.

> What I would like is an array with just the values of the daily averages .
> Additional a report-like array output with the format
> day value
> 1   3
> 2   11
>
> would be nice.

Mmh, contact me offlist, I have no idea what you're trying to do here.


> Thanks again for the ice package. Things are looking nice!

thanks a lot for using it, that's the best way to improve it !


From fie.pye at atlas.cz  Thu Dec 20 13:12:46 2007
From: fie.pye at atlas.cz (Pye Fie)
Date: Thu, 20 Dec 2007 18:12:46 GMT
Subject: [SciPy-user] MayaVi 2 instalation
Message-ID: <2ca156948383499cb604e64687b32ea0@db3381b8eaa54a2faa19d0073d9697e1>

Hello.

I have got :
PC: 2x Dual-core Opteron 275
OS: CentOS 5.0, kernel 2.6.18-8.1.15.el5

I would like to prepare a computational environment based on Python and Python modules. I have already compiled and installed Python2.5.1 and modules such as Ipython, SciPy, NumPy, tables, mpi4py, matplotlib, wxpython and  so on, libraries such as BLAS, LAPACK, ATLAS, HDF5, MPICH, VTK.  Now I would like to continue with compilation and installation of MayaVi 2. 

According to instructions in
https://svn.enthought.com/enthought/wiki/GrabbingAndBuilding
I should install:
Setuptools
Traits
TVTK
MayaVi 2

As I  understand the instructions in Setuptools README, Setuptools will install a new Python and new modules. But I don't need a new Python installation. I have one in /usr/local/python/python2.5.1 directory. 

I studied SciPy ML discussion 
http://projects.scipy.org/pipermail/scipy-user/2007-November/014374.html
and downloaded files from
http://code.enthought.com/downloads/source/ets2.6/
with intention that it helps. Unfortunately it also requires Setuptools.

1. Is there a solution to my problem?
or 
2. Am I wrong in understanding installation instruction?

Could anybody help me?
Best regards.
fie pie

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

http://search.atlas.cz/

From robert.kern at gmail.com  Thu Dec 20 16:25:47 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 20 Dec 2007 15:25:47 -0600
Subject: [SciPy-user] MayaVi 2 instalation
In-Reply-To: <2ca156948383499cb604e64687b32ea0@db3381b8eaa54a2faa19d0073d9697e1>
References: <2ca156948383499cb604e64687b32ea0@db3381b8eaa54a2faa19d0073d9697e1>
Message-ID: <476ADDDB.30100@gmail.com>

Pye Fie wrote:
> Hello.
> 
> I have got :
> PC: 2x Dual-core Opteron 275
> OS: CentOS 5.0, kernel 2.6.18-8.1.15.el5
> 
> I would like to prepare a computational environment based on Python and Python modules. I have already compiled and installed Python2.5.1 and modules such as Ipython, SciPy, NumPy, tables, mpi4py, matplotlib, wxpython and  so on, libraries such as BLAS, LAPACK, ATLAS, HDF5, MPICH, VTK.  Now I would like to continue with compilation and installation of MayaVi 2. 

Hi, you will want to join us on enthought-dev rather than scipy-user.

  https://mail.enthought.com/mailman/listinfo/enthought-dev

> According to instructions in
> https://svn.enthought.com/enthought/wiki/GrabbingAndBuilding
> I should install:
> Setuptools
> Traits
> TVTK
> MayaVi 2
> 
> As I  understand the instructions in Setuptools README, Setuptools will install a new Python and new modules.

This is incorrect. Just follow the instructions here:

  http://pypi.python.org/pypi/setuptools

-- 
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 dwf at cs.toronto.edu  Thu Dec 20 17:15:47 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 20 Dec 2007 17:15:47 -0500
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <4769CA5C.70205@enthought.com>
References: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
	<4769CA5C.70205@enthought.com>
Message-ID: <7047671E-184F-4160-9562-027C68C25948@cs.toronto.edu>


On 19-Dec-07, at 8:50 PM, Travis E. Oliphant wrote:

> Yes, absolutely, warnings should be there in the docstring.  Your
> perspective is not illogical nor are many of us unsympathetic to it.
> But,  following your rule would be an "exceptional" case  to an
> otherwise easy to explain rule.
>
> Our approach thus far has been to educate users coming from other
> languages that there is such a thing as a 1D array that is neither a
> "row" nor a "column" vector and so in fact a transpose operation is  
> not
> really what you meant (as it does nothing to a 1D array).   Instead  
> what
> you wanted to do was convert the 1-D array to a 2-D array with 1  
> row and
> then transpose that to a 2D array with 1 column.

Would it be totally crazy to throw an exception if you try to  
access .T on a 1D array?  Since it's not meaningful, I don't see a  
compelling reason why it should need to succeed.

Also, would more... aggressively encouraging use of the "matrix" type  
help alleviate the migration pains?

David


From matthieu.brucher at gmail.com  Thu Dec 20 17:21:56 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 20 Dec 2007 23:21:56 +0100
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <7047671E-184F-4160-9562-027C68C25948@cs.toronto.edu>
References: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
	<4769CA5C.70205@enthought.com>
	<7047671E-184F-4160-9562-027C68C25948@cs.toronto.edu>
Message-ID: <e76aa17f0712201421q888a919v528e7c992a4e7071@mail.gmail.com>

>
> Would it be totally crazy to throw an exception if you try to
> access .T on a 1D array?  Since it's not meaningful, I don't see a
> compelling reason why it should need to succeed.
>

A really bad idea, I think. There are some cases where you might want to do
this in generic code, so this will only be slower and cumbersome for people
that want to do things correctly. It IS meaningful.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071220/f336825a/attachment.html>

From dwf at cs.toronto.edu  Thu Dec 20 17:39:30 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 20 Dec 2007 17:39:30 -0500
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <e76aa17f0712201421q888a919v528e7c992a4e7071@mail.gmail.com>
References: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
	<4769CA5C.70205@enthought.com>
	<7047671E-184F-4160-9562-027C68C25948@cs.toronto.edu>
	<e76aa17f0712201421q888a919v528e7c992a4e7071@mail.gmail.com>
Message-ID: <B33F952E-75DE-418D-A062-09C5DC6E0A25@cs.toronto.edu>


On 20-Dec-07, at 5:21 PM, Matthieu Brucher wrote:
>>
>> Would it be totally crazy to throw an exception if you try to
>> access .T on a 1D array?  Since it's not meaningful, I don't see a
>> compelling reason why it should need to succeed.
>>

I can't think of a situation where code would need to be generic  
enough to need to acceptably call what's basically a no-op on a 1D  
array. But you might be right. Still, I think it should be  
discouraged somehow, for the benefit of new users; perhaps a warn()?

David


From dwf at cs.toronto.edu  Thu Dec 20 17:40:48 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 20 Dec 2007 17:40:48 -0500
Subject: [SciPy-user] arraysetops & string arrays
In-Reply-To: <476A448D.3040206@ntc.zcu.cz>
References: <8C1A7051-36CA-4B8E-A1F0-0E4C6834BD7C@cs.toronto.edu>
	<476A448D.3040206@ntc.zcu.cz>
Message-ID: <2300E41E-DBF2-485D-9159-EC602A6BCC22@cs.toronto.edu>

On 20-Dec-07, at 5:31 AM, Robert Cimrman wrote:
>
> Fixed in SVN. It was more a missing feature than a bug - I did not
> realized that it should work for non-numerical arrays, too.
>
> If you do not use SVN, I can mail you the fix, just tell me.

Wonderful! I haven't been using the bleeding edge SVN version of  
numpy yet, but I can grab just this file, no need to mail it. Thanks  
very much!

David



From matthieu.brucher at gmail.com  Fri Dec 21 01:56:13 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 21 Dec 2007 07:56:13 +0100
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <B33F952E-75DE-418D-A062-09C5DC6E0A25@cs.toronto.edu>
References: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
	<4769CA5C.70205@enthought.com>
	<7047671E-184F-4160-9562-027C68C25948@cs.toronto.edu>
	<e76aa17f0712201421q888a919v528e7c992a4e7071@mail.gmail.com>
	<B33F952E-75DE-418D-A062-09C5DC6E0A25@cs.toronto.edu>
Message-ID: <e76aa17f0712202256l7b684726mf00cde8ac6f2fe9@mail.gmail.com>

>
> I can't think of a situation where code would need to be generic
> enough to need to acceptably call what's basically a no-op on a 1D
> array. But you might be right. Still, I think it should be
> discouraged somehow, for the benefit of new users; perhaps a warn()?
>

This remains the same : why do you want to change something that is correct
to something that is not ? Why do you want to put a warning when people want
to do generic code ?
I do a lot of generic code, and the fact that .T does nothing for a 1D array
and does not put a warning is the behaviour I expect from Numpy. The
behaviour in Matlab is not the correct one, I can't count the number of
times I had to explain to a student why what he does does not give the
expected result.
Python is about to make things simple and clear, let's jus do not the
opposite in Numpy.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071221/0a1efea0/attachment.html>

From cohen at slac.stanford.edu  Fri Dec 21 12:07:30 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Fri, 21 Dec 2007 09:07:30 -0800
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <e76aa17f0712202256l7b684726mf00cde8ac6f2fe9@mail.gmail.com>
References: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>	<4769CA5C.70205@enthought.com>	<7047671E-184F-4160-9562-027C68C25948@cs.toronto.edu>	<e76aa17f0712201421q888a919v528e7c992a4e7071@mail.gmail.com>	<B33F952E-75DE-418D-A062-09C5DC6E0A25@cs.toronto.edu>
	<e76aa17f0712202256l7b684726mf00cde8ac6f2fe9@mail.gmail.com>
Message-ID: <476BF2D2.8050406@slac.stanford.edu>

IMHO I think that we are reaching noise level on this thread.... Erik 
and Travis's option to put a warning in the doc string seems reasonable, 
easy, and efficient to me. numpy.array can but is not meant to do LinA 
lg matrix computation, acknowledged! That should be CAPITALIZED in 
tutorials, docstrings, whathever, and that should be enough.
cheers,
Johann

Matthieu Brucher wrote:
>
>     I can't think of a situation where code would need to be generic
>     enough to need to acceptably call what's basically a no-op on a 1D
>     array. But you might be right. Still, I think it should be
>     discouraged somehow, for the benefit of new users; perhaps a warn()?
>
>
> This remains the same : why do you want to change something that is 
> correct to something that is not ? Why do you want to put a warning 
> when people want to do generic code ?
> I do a lot of generic code, and the fact that .T does nothing for a 1D 
> array and does not put a warning is the behaviour I expect from Numpy. 
> The behaviour in Matlab is not the correct one, I can't count the 
> number of times I had to explain to a student why what he does does 
> not give the expected result.
> Python is about to make things simple and clear, let's jus do not the 
> opposite in Numpy.
>
> Matthieu
> -- 
> French PhD student
> Website : http://matthieu-brucher.developpez.com/ 
> <http://matthieu-brucher.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
>   



From aisaac at american.edu  Fri Dec 21 12:34:00 2007
From: aisaac at american.edu (Alan G Isaac)
Date: Fri, 21 Dec 2007 12:34:00 -0500
Subject: [SciPy-user] Mathematica Element-wise Multiplication
In-Reply-To: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
References: <a47088770712191632m261b50br17c885f48ab6e916@mail.gmail.com>
Message-ID: <Mahogany-0.67.0-1280-20071221-123400.00@american.edu>

On Wed, 19 Dec 2007, Erik Volz apparently wrote:
> z+=z.T # where dimension z > 1 
> which does not do what a naive user would think it does (try it). 

You mean like this? ::

    >>> x=N.ones( (2,2) )
    >>> x
    array([[ 1.,  1.],
           [ 1.,  1.]])
    >>> x += x.T
    >>> x
    array([[ 2.,  2.],
           [ 3.,  2.]])

That just shows you that the in-place updating takes place 
in row major order.  You have to have something like this
or create temporary arrays, but not creating temporary
arrays is a core value of the in-place operators.

If you do not like them, do not use them.
Or force the creation of a temporary array on the RHS::

     >>> x=N.ones( (2,2) )
     >>> x += x.T + 0
     >>> x
     array([[ 2.,  2.],
            [ 2.,  2.]])

Cheers,
Alan Isaac





From nikolaskaralis at gmail.com  Sun Dec 23 11:37:03 2007
From: nikolaskaralis at gmail.com (Nikolas Karalis)
Date: Sun, 23 Dec 2007 18:37:03 +0200
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <db6b5ecc0712111205s1f78dce2tfdf3c501db754d78@mail.gmail.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>
	<475EAEC1.4020501@enthought.com>
	<85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>
	<e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>
	<db6b5ecc0712111205s1f78dce2tfdf3c501db754d78@mail.gmail.com>
Message-ID: <85b2e0230712230837h2356c230h431af5f9c283fbdf@mail.gmail.com>

Hello again. I return with another question regarding characteristic
polynomials.

I am computing the characteristic polynomials for matrices. And while (for
millions of them) i get the right result, i found some cases, where i get
the "wrong".
Let me give you the examples, and then i will explain.

>>> A=matrix([[290, 324, 323, 364, 340, 341, 365, 336, 342, 326],
        [324, 290, 322, 338, 366, 341, 365, 336, 344, 324],
        [323, 322, 286, 337, 337, 366, 364, 361, 320, 323],
        [345, 327, 326, 315, 343, 344, 342, 363, 343, 312],
        [329, 347, 325, 343, 319, 345, 343, 364, 327, 327],
        [329, 329, 347, 344, 345, 323, 344, 342, 348, 328],
        [347, 347, 346, 342, 343, 344, 321, 341, 326, 311],
        [325, 325, 343, 363, 364, 342, 341, 313, 324, 309],
        [342, 344, 320, 362, 338, 366, 338, 335, 286, 307],
        [340, 339, 338, 330, 361, 362, 333, 329, 316, 265]])

>>> B=matrix([[290, 324, 336, 365, 341, 340, 364, 323, 342, 326],
        [324, 290, 336, 339, 367, 340, 364, 322, 344, 324],
        [325, 325, 313, 341, 342, 364, 363, 343, 324, 309],
        [346, 328, 341, 319, 346, 343, 342, 347, 345, 312],
        [330, 348, 342, 346, 321, 345, 344, 346, 329, 327],
        [328, 328, 364, 343, 345, 321, 341, 326, 346, 328],
        [346, 346, 363, 342, 344, 341, 317, 325, 324, 311],
        [323, 322, 361, 365, 365, 338, 336, 286, 320, 323],
        [342, 344, 335, 364, 340, 364, 336, 320, 286, 307],
        [340, 339, 329, 332, 363, 360, 331, 338, 316, 265]])

>>> poly(A)
array([  1.00000000e+00,  -3.00800000e+03,  -1.10612600e+06,
        -1.55679754e+08,  -1.10343405e+10,  -4.15092661e+11,
        -7.89507268e+12,  -6.51631023e+13,  -1.80794492e+14,
        -2.21111889e+00,  -2.95926167e-13])

>>> poly(B)
array([  1.00000000e+00,  -3.00800000e+03,  -1.10612600e+06,
        -1.55679754e+08,  -1.10343405e+10,  -4.15092661e+11,
        -7.89507268e+12,  -6.51631023e+13,  -1.80794492e+14,
         4.59224482e+00,   3.39308585e-14])

As you can see, the 2 polynomials are the same (up to the last 2 terms). The
last term is considered to be "almost" 0, so we can see that the difference
is the coefficient of x^1.

If we compute the exact same thing with Maple and Sage, we get (for both
matrices) :

*x^10 - 3008*x^9 - 1106126*x^8 - 155679754*x^7 - 11034340454*x^6 -
415092661064*x^5 - 7895072675601*x^4 - 65163102265268*x^3 -
180794492489124*x^2*

so, it is the same, since it doesn't "compute" the x^1 term.

This also happens for a few other matrices...

Can anybody help me with this? What is the right answer for the char. poly
(i guess it is the Sage's and Maple's one, since they agree).
Why does this difference occur? Is it "sage" to ignore the last 2 terms?

Thank you in advance.

Nikolas

On Dec 11, 2007 10:05 PM, Fernando Perez <fperez.net at gmail.com> wrote:

> On Dec 11, 2007 9:45 AM, Matthieu Brucher <matthieu.brucher at gmail.com>
> wrote:
> > You can't expect scipy to find the exact coefficients. sage is a
> symbolic
> > package, it will find the correct answers, but scipy will find only an
> > approximated one, up to the machine precision. This is what you see in
> your
> > exemple.
> > If you have integers, you could expect scipy to return long integers
> (exact
> > result), but this is not the case as almost everything is converted into
> a
> > float array before the actual C or Fortran routine is run.
>
> In this case Sage isn't using anything symbolic though: the issue is
> that Sage has from the ground up defined a type system where it knows
> what field its inputs are being computed over.  So if a matrix is made
> up of only integers, it knows that computations are to be performed in
> exact arithmetic over Z, and does so accordingly.
>
> Sage's origins are actually number theory, not symbolic computing, and
> its strongest area is precisely in exact arithmetic, with enormous
> sophistication for some of its algorithms.  Initially its (free)
> symbolic capabilities were all obtained via calls to Maxima, though
> now that Ondrej is actively helping with SymPy integration into Sage,
> that is changing and SymPy is natively available as well, which means
> that over time there will be more and more native (python) symbolic
> support as well.
>
> I'd agree though that for this kind of calculation, Sage is the tool to
> use.
>
> Cheers,
>
> f
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
Nikolas Karalis
Applied Mathematics and Physics Undergraduate
National Technical University of Athens, Greece
http://users.ntua.gr/ge04042
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071223/f0e4d06c/attachment.html>

From peridot.faceted at gmail.com  Sun Dec 23 13:53:09 2007
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Sun, 23 Dec 2007 13:53:09 -0500
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <85b2e0230712230837h2356c230h431af5f9c283fbdf@mail.gmail.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>
	<475EAEC1.4020501@enthought.com>
	<85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>
	<e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>
	<db6b5ecc0712111205s1f78dce2tfdf3c501db754d78@mail.gmail.com>
	<85b2e0230712230837h2356c230h431af5f9c283fbdf@mail.gmail.com>
Message-ID: <ce557a360712231053n7638c806ycd081bf8fe5c18a1@mail.gmail.com>

On 23/12/2007, Nikolas Karalis <nikolaskaralis at gmail.com> wrote:

> >>> A=matrix([[290, 324, 323, 364, 340, 341, 365, 336, 342, 326],
>         [324, 290, 322, 338, 366, 341, 365, 336, 344, 324],
>         [323, 322, 286, 337, 337, 366, 364, 361, 320, 323],
>         [345, 327, 326, 315, 343, 344, 342, 363, 343, 312],
>         [329, 347, 325, 343, 319, 345, 343, 364, 327, 327],
>         [329, 329, 347, 344, 345, 323, 344, 342, 348, 328],
>         [347, 347, 346, 342, 343, 344, 321, 341, 326, 311],
>         [325, 325, 343, 363, 364, 342, 341, 313, 324, 309],
>         [342, 344, 320, 362, 338, 366, 338, 335, 286, 307],
>         [340, 339, 338, 330, 361, 362, 333, 329, 316, 265]])
>
> >>> B=matrix([[290, 324, 336, 365, 341, 340, 364, 323, 342, 326],
>         [324, 290, 336, 339, 367, 340, 364, 322, 344, 324],
>         [325, 325, 313, 341, 342, 364, 363, 343, 324, 309],
>         [346, 328, 341, 319, 346, 343, 342, 347, 345, 312],
>         [330, 348, 342, 346, 321, 345, 344, 346, 329, 327],
>         [328, 328, 364, 343, 345, 321, 341, 326, 346, 328],
>         [346, 346, 363, 342, 344, 341, 317, 325, 324, 311],
>         [323, 322, 361, 365, 365, 338, 336, 286, 320, 323],
>         [342, 344, 335, 364, 340, 364, 336, 320, 286, 307],
>         [340, 339, 329, 332, 363, 360, 331, 338, 316, 265]])
>
> >>> poly(A)
> array([  1.00000000e+00,  -3.00800000e+03,  -1.10612600e+06,
>         -1.55679754e+08,  -1.10343405e+10,  -4.15092661e+11,
>          -7.89507268e+12,  -6.51631023e+13,  -1.80794492e+14,
>         -2.21111889e+00,  -2.95926167e-13])
>
> >>> poly(B)
> array([  1.00000000e+00,  -3.00800000e+03,  -1.10612600e+06,
>         -1.55679754e+08 ,  -1.10343405e+10,  -4.15092661e+11,
>         -7.89507268e+12,  -6.51631023e+13,  -1.80794492e+14,
>          4.59224482e+00,   3.39308585e-14])
>
> As you can see, the 2 polynomials are the same (up to the last 2 terms). The
> last term is considered to be "almost" 0, so we can see that the difference
> is the coefficient of x^1.
>
> If we compute the exact same thing with Maple and Sage, we get (for both
> matrices) :
> x^10 - 3008*x^9 - 1106126*x^8 - 155679754*x^7 - 11034340454*x^6 -
> 415092661064*x^5 - 7895072675601*x^4 - 65163102265268*x^3 -
> 180794492489124*x^2
>
> so, it is the same, since it doesn't "compute" the x^1 term.
>
> This also happens for a few other matrices...
>
> Can anybody help me with this? What is the right answer for the char. poly
> (i guess it is the Sage's and Maple's one, since they agree).
> Why does this difference occur? Is it "sage" to ignore the last 2 terms?

When you do all but the simplest matrix computations with numpy, it
treats the matrices as  containing floating-point numbers. Since
floating-point numbers are as a rule only approximations to the exact
value you would like ot be working with, the answers you get out are
only approximations. If you look at the coefficients you're getting
from numpy's characteristic polynomial calculation, you will see that
the coefficient of x^1 is about 10^-14 times the coefficient of x^2. A
good rule of thumb is that you can consider as almost zero any number
that is many orders of magnitude less than the biggest number in your
answer. So 4, or 2, or whatever, is "close to zero" in this context.

More specifically, when you compute a small number as the difference
of two very large numbers, that number has far fewer digits of
accuracy than the original two numbers. Many matrix computations, and
in particular eigenvalue and characteristic polynomial calculations,
involve just this kind of delicate cancellation, often many times
over. So you have to expect errors to accumulate.

Some problems, like finding the roots of a polynomial whose
coefficients you know, are just inherently "unstable", that is, if you
change the coefficients by one in the last significant figure, the
roots move around wildly. These problems are basically hopeless, and
you will need to reformulate your problem in a way that doesn't
require you to caluclate the roots. Matrix inversion can be like this,
for example, so rather than solve Ax=b as x=A^-1 b, you should use a
specialized solver (of which there are many). For your problem, the
answer depends on what you are doing with the characteristic
polynomials. If you are (for example) trying to find all the
eigenvalues, use numpy's eigenvalue solvers directly. If you only want
certain coefficients - the x^0 and x^(n-1) are the easiest - think
about whether there's a better formula (e.g. determinant or trace) for
them.

Your case is a bit special: your input numbers are not floating-point.
They are integers, and (I assume) they represent the input values
*exactly*. Thus in principle it should be possible to compute the
characteristic polynomial exactly. Unfortunately, using floating-point
is almost certainly a bad way to do this. (MAPLE might allow you to
use floating-point with (say) several hundred decimal places, which
might result in very nearly integral coefficients, but this is going
to be *slow*.) Instead, you should use a special-purpose integer
linear algebra routine. These integer linear algebra routines are
specialized, computationally expensive, and not part of numpy. They
are part of SAGE and MAPLE. More generally, this kind of exact
computation belongs to the domain of "symbolic computation", which is
not something numpy is for.

In summary: use SAGE, or think about your problem more.

Anne


From robert.kern at gmail.com  Sun Dec 23 13:42:30 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Sun, 23 Dec 2007 12:42:30 -0600
Subject: [SciPy-user] Characteristic polynomial
In-Reply-To: <85b2e0230712230837h2356c230h431af5f9c283fbdf@mail.gmail.com>
References: <85b2e0230712110434x3b71be92sa2607f75886410ed@mail.gmail.com>	<475EAEC1.4020501@enthought.com>	<85b2e0230712110838n19e0e229m73f80d1a61fd3cc3@mail.gmail.com>	<e76aa17f0712110845s69b42f64l8f8039029731e52e@mail.gmail.com>	<db6b5ecc0712111205s1f78dce2tfdf3c501db754d78@mail.gmail.com>
	<85b2e0230712230837h2356c230h431af5f9c283fbdf@mail.gmail.com>
Message-ID: <476EAC16.1040908@gmail.com>

Nikolas Karalis wrote:
> Hello again. I return with another question regarding characteristic
> polynomials.
> 
> I am computing the characteristic polynomials for matrices. And while
> (for millions of them) i get the right result, i found some cases, where
> i get the "wrong".
> Let me give you the examples, and then i will explain.
> 
>>>> A=matrix([[290, 324, 323, 364, 340, 341, 365, 336, 342, 326],
>         [324, 290, 322, 338, 366, 341, 365, 336, 344, 324],
>         [323, 322, 286, 337, 337, 366, 364, 361, 320, 323],
>         [345, 327, 326, 315, 343, 344, 342, 363, 343, 312],
>         [329, 347, 325, 343, 319, 345, 343, 364, 327, 327],
>         [329, 329, 347, 344, 345, 323, 344, 342, 348, 328],
>         [347, 347, 346, 342, 343, 344, 321, 341, 326, 311],
>         [325, 325, 343, 363, 364, 342, 341, 313, 324, 309],
>         [342, 344, 320, 362, 338, 366, 338, 335, 286, 307],
>         [340, 339, 338, 330, 361, 362, 333, 329, 316, 265]])
> 
>>>> B=matrix([[290, 324, 336, 365, 341, 340, 364, 323, 342, 326],
>         [324, 290, 336, 339, 367, 340, 364, 322, 344, 324],
>         [325, 325, 313, 341, 342, 364, 363, 343, 324, 309],
>         [346, 328, 341, 319, 346, 343, 342, 347, 345, 312],
>         [330, 348, 342, 346, 321, 345, 344, 346, 329, 327],
>         [328, 328, 364, 343, 345, 321, 341, 326, 346, 328],
>         [346, 346, 363, 342, 344, 341, 317, 325, 324, 311],
>         [323, 322, 361, 365, 365, 338, 336, 286, 320, 323],
>         [342, 344, 335, 364, 340, 364, 336, 320, 286, 307],
>         [340, 339, 329, 332, 363, 360, 331, 338, 316, 265]])
> 
>>>> poly(A)
> array([  1.00000000e+00,  -3.00800000e+03,  -1.10612600e+06,
>         -1.55679754e+08,  -1.10343405e+10,  -4.15092661e+11,
>         -7.89507268e+12,  -6.51631023e+13,  -1.80794492e+14,
>         -2.21111889e+00,  -2.95926167e-13])
> 
>>>> poly(B)
> array([  1.00000000e+00,  -3.00800000e+03,  -1.10612600e+06,
>         -1.55679754e+08 ,  -1.10343405e+10,  -4.15092661e+11,
>         -7.89507268e+12,  -6.51631023e+13,  -1.80794492e+14,
>          4.59224482e+00,   3.39308585e-14])
> 
> As you can see, the 2 polynomials are the same (up to the last 2 terms).
> The last term is considered to be "almost" 0, so we can see that the
> difference is the coefficient of x^1.
> 
> If we compute the exact same thing with Maple and Sage, we get (for both
> matrices) :
> 
> *x^10 - 3008*x^9 - 1106126*x^8 - 155679754*x^7 - 11034340454*x^6 - 415092661064*x^5 - 7895072675601*x^4 - 65163102265268*x^3 - 180794492489124*x^2
> *
> 
> so, it is the same, since it doesn't "compute" the x^1 term.
> 
> This also happens for a few other matrices...
> 
> Can anybody help me with this? What is the right answer for the char.
> poly (i guess it is the Sage's and Maple's one, since they agree).

Yes, also because they are using exact arithmetic rather than floating point.

> Why does this difference occur? Is it "sage" to ignore the last 2 terms?

Remember that floating point errors are usually relative, not absolute. A value
of 1e-14 is not always "almost" 0, and a value of 1e0 can sometimes be "almost"
0 when the other values are around 1e+14. It should make you suspicious that the
exponent steadily increases then suddenly drops to 0.

You can do a little bit of cleanup on the floating-point results by looking at
the eigenvalues directly.

In [14]: ev = linalg.eigvals(A)

In [15]: ev
Out[15]:
array([  3.35212801e+03 +0.00000000e+00j,
        -9.14584932e+01 +0.00000000e+00j,
        -7.67326852e+01 +0.00000000e+00j,
        -6.72710568e+01 +0.00000000e+00j,
        -5.92400278e+01 +0.00000000e+00j,
        -3.36474895e+01 +0.00000000e+00j,
        -1.01081162e+01 +0.00000000e+00j,
        -5.67013979e+00 +0.00000000e+00j,
         9.77724356e-15 +8.22960451e-15j,   9.77724356e-15 -8.22960451e-15j])


You can see that the last two terms are "almost" 0 in the relative sense.
Therefore, you can set them to zero before you construct the polynomial:

In [24]: aev = absolute(ev)

In [25]: ev[(aev / aev.max()) < 1e-16] = 0.0

In [26]: poly(real_if_close(ev))
Out[26]:
array([  1.00000000e+00,  -3.00800000e+03,  -1.10612600e+06,
        -1.55679754e+08,  -1.10343405e+10,  -4.15092661e+11,
        -7.89507268e+12,  -6.51631023e+13,  -1.80794492e+14,
         0.00000000e+00,   0.00000000e+00])

But if you want really reliable results for largish integer matrices, you'll
want to use SAGE. Your matrices are large enough that the floating point error
will be significant even if there are no "almost" 0s to clean 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 ac1201 at gmail.com  Sun Dec 23 19:42:21 2007
From: ac1201 at gmail.com (Andrew Charles)
Date: Mon, 24 Dec 2007 11:42:21 +1100
Subject: [SciPy-user] reading data into a RecordArray
Message-ID: <b375ebf50712231642x6ab5cf00kbb63a16435ac61ee@mail.gmail.com>

I'm relatively new to python, and new to this list. Hello everyone - I
really like what you're doing with scipy.

My question relates to code I'm writing to convert my ASCII data to netCDF.
I'm trying to:

1. Read in the data file
2. Read in a file containing the column headers
3. Merge the two into a record array

I've read the overview (http://www.scipy.org/RecordArrays), and gone back
over the introductory material on lists and tuples, but I've stumped myself.
The code

    from scipy import *

    colfile = open("my_column_spec.txt")
    columns=[]
    cols = colfile.readlines()
    for col in cols:
        columns = columns + [(col,'f4')]

    ifile = open("my_ascii_data.txt")
    data = io.array_import.read_array(ifile)

    data = array(data,columns)

fails on the last line with the error "expected a readable buffer object".
In case it's relevant, dt.shape is (1984,22) and len(cols) is 22. Any
pointers as to what I'm doing wrong?


Cheers,

Andrew Charles
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071224/d601d74b/attachment.html>

From fperez.net at gmail.com  Mon Dec 24 01:02:48 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Sun, 23 Dec 2007 23:02:48 -0700
Subject: [SciPy-user] reading data into a RecordArray
In-Reply-To: <b375ebf50712231642x6ab5cf00kbb63a16435ac61ee@mail.gmail.com>
References: <b375ebf50712231642x6ab5cf00kbb63a16435ac61ee@mail.gmail.com>
Message-ID: <db6b5ecc0712232202i333f1fd2nd7d502a1e55d1c74@mail.gmail.com>

On Dec 23, 2007 5:42 PM, Andrew Charles <ac1201 at gmail.com> wrote:
> I'm relatively new to python, and new to this list. Hello everyone - I
> really like what you're doing with scipy.
>
> My question relates to code I'm writing to convert my ASCII data to netCDF.
> I'm trying to:
>
> 1. Read in the data file
> 2. Read in a file containing the column headers
> 3. Merge the two into a record array
>
> I've read the overview (http://www.scipy.org/RecordArrays ), and gone back
> over the introductory material on lists and tuples, but I've stumped myself.
> The code
>
>     from scipy import *
>
>     colfile = open("my_column_spec.txt")
>     columns=[]
>      cols = colfile.readlines()
>     for col in cols:
>         columns = columns + [(col,'f4')]

Try at this point in the code:

columns = tuple(columns)

>
>     ifile = open("my_ascii_data.txt")
>     data = io.array_import.read_array(ifile)
>
>     data = array(data,columns)
>
> fails on the last line with the error "expected a readable buffer object".
> In case it's relevant, dt.shape is (1984,22) and len(cols) is 22. Any
> pointers as to what I'm doing wrong?

And let us know if it helps.

Cheers,

f


From fperez.net at gmail.com  Mon Dec 24 16:43:51 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Mon, 24 Dec 2007 14:43:51 -0700
Subject: [SciPy-user] From the Sage/Jmol lists: "scholarly activity"?
Message-ID: <db6b5ecc0712241343p955f12dod38b9c8801417207@mail.gmail.com>

Here's something that came up on the jmol list and that William Stein
(SAGE lead) forwarded, but which is also a concern to this list.

Sage has also had discussions on the matter and the idea of a Sage
journal (http://sagemath.org/jsage/) exists already, in a similar vein
to previous discussions here on a 'scipy journal'.

I don't have any answer to this at the moment, but I'm pretty sure
it's something that will recur.  I personally know of several people
who have been bitten by this very same problem.  It's worth noting
that the original author of the JMol message is a *tenured* faculty
member at U. Wisconsin, so not even tenure seems to be a complete
protection against the perception from many academic administrators
that contributing to open source scientific projects is "not scholarly
work".

Cheers,

f

---------- Forwarded message ----------
From: William Stein <wstein at gmail.com>
Date: Dec 24, 2007 12:55 PM
Subject: [sage-devel] "scholarly activity"?
To: sage-devel <sage-devel at googlegroups.com>



Hi,

I just noticed this email on the jmol developer mailing list.  See below.

if anybody has any thoughts or ideas -- long or short term -- about how to
structure or restructure sage development so the same sort of thing doesn't
happen to us, please speak up.   I think something like JSAGE
(http://sagemath.org/jsage/)
-- if it were to take off -- would really help.

http://sourceforge.net/mailarchive/forum.php?thread_name=F4FB859D-42D1-4974-B244-E67DE7F3681F%40uwosh.edu&forum_name=jmol-developers

"
Dear Jmol team:
This letter is to notify you that I will not be able to continue
participating in Jmol development. I hope this situation will only
be temporary. However, my rejoining the project depends on my
institution being convinced to give me some kind of credit for these
activities. Since my contributions to the Jmol project were not
deemed a "form of off-campus, peer reviewed scholarly or artistic
product", I cannot afford to put any more time into the project. As
it stands, I have been told that my scholarly activity level has not
been adequate recently. To help me increase my scholarly output, I
will have to teach 3-6 hours more of classes each week, and will be
responsible for grading additional work from 20-60 more students.
Somehow that is supposed to increase my scholarly activity:) Put
simply, this means I will not have time to devote to Jmol or most of
my other scholarly activities. Since Jmol is the scholarly activity
that doesn't presently count, I need to drop that and use any time I
can eke out for scholarship of other kinds. If I can convince my
department to include contributions to projects like Jmol on a list
of creditable activity, I will be able to rejoin the project. This
is likely to take a semester. Thus I hope to be able to rejoin the
project next spring, about May.

I will stay on the lists and try to stay up-to-date. I also plan to
spend some of the holiday break bringing the Wiki up-to-date on the
export to web function. I think it will help my case if export to
web is better documented and thus gets used more.

I have found my involvement in the project intellectually
stimulating and am sorry that my situation requires that I halt my
participation.

Sincerely,
Jonathan
Dr. Jonathan H. Gutow
Chemistry Department gutow at uw...
UW-Oshkosh Office:920-424-1326
800 Algoma Boulevard FAX:920-424-2042
Oshkosh, WI 54901
http://www.uwosh.edu/faculty_staff/gutow/ "


A response:

This is quite unfortunate, Jonathan. Please forward my regrets to your
Department Chair.
Contributions like yours are hard to come by, and, for heaven's sakes,
if this isn't peer reviewed,
WHAT IS? I think it's just a sign that people have not caught up with
the times. Very unfortunate.

Bob Hanson
Professor of Chemistry
St. Olaf College


From gael.varoquaux at normalesup.org  Mon Dec 24 16:57:01 2007
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 24 Dec 2007 22:57:01 +0100
Subject: [SciPy-user] From the Sage/Jmol lists: "scholarly activity"?
In-Reply-To: <db6b5ecc0712241343p955f12dod38b9c8801417207@mail.gmail.com>
References: <db6b5ecc0712241343p955f12dod38b9c8801417207@mail.gmail.com>
Message-ID: <20071224215701.GD8488@clipper.ens.fr>

On Mon, Dec 24, 2007 at 02:43:51PM -0700, Fernando Perez wrote:
> Here's something that came up on the jmol list and that William Stein
> (SAGE lead) forwarded, but which is also a concern to this list.

> Sage has also had discussions on the matter and the idea of a Sage
> journal (http://sagemath.org/jsage/) exists already, in a similar vein
> to previous discussions here on a 'scipy journal'.

> I don't have any answer to this at the moment, but I'm pretty sure
> it's something that will recur.  I personally know of several people
> who have been bitten by this very same problem.  It's worth noting
> that the original author of the JMol message is a *tenured* faculty
> member at U. Wisconsin, so not even tenure seems to be a complete
> protection against the perception from many academic administrators
> that contributing to open source scientific projects is "not scholarly
> work".

Yes, it is a very important problem. Indeed it would be nice to have a
way of get scholar-like credit for work on open source projects. If this
means creating a journal dedicated to released scientific code (let us
just avoid putting "open source" in the title, it's too political), than
it may be the way to go. I think it should not be restricted to Python
and friends, a broader could increase its visibility.

Any ideas welcome, there is a Damocles sword hanging above many
contributions.

Ga?l


From william.ratcliff at gmail.com  Mon Dec 24 17:30:24 2007
From: william.ratcliff at gmail.com (william ratcliff)
Date: Mon, 24 Dec 2007 17:30:24 -0500
Subject: [SciPy-user] From the Sage/Jmol lists: "scholarly activity"?
In-Reply-To: <20071224215701.GD8488@clipper.ens.fr>
References: <db6b5ecc0712241343p955f12dod38b9c8801417207@mail.gmail.com>
	<20071224215701.GD8488@clipper.ens.fr>
Message-ID: <827183970712241430s2b907f79t951205a0b8a4b730@mail.gmail.com>

At my workplace a very successful gui for use in diffraction was developed
called EXPGUI.  They published a paper describing it and asked people who
used it to cite it.  It is one of the most highly cited papers of our
facility.  For some utilities such as jmol (or matplotlib), could that
help?   Another approach that I have seen used is to require a registration
(free of course!) for people who want to download the  program (or tracking
of download statistics) to show impact to one's supervisors....Another
question--how would his department have looked upon an E-journal?  Another
question might be is whether or not a funding agency such as the NSF could
offer grants for work on OSS so that faculty who work on it could get more
credit from their departments?

Cheers,
William

On Dec 24, 2007 4:57 PM, Gael Varoquaux <gael.varoquaux at normalesup.org>
wrote:

> On Mon, Dec 24, 2007 at 02:43:51PM -0700, Fernando Perez wrote:
> > Here's something that came up on the jmol list and that William Stein
> > (SAGE lead) forwarded, but which is also a concern to this list.
>
> > Sage has also had discussions on the matter and the idea of a Sage
> > journal (http://sagemath.org/jsage/) exists already, in a similar vein
> > to previous discussions here on a 'scipy journal'.
>
> > I don't have any answer to this at the moment, but I'm pretty sure
> > it's something that will recur.  I personally know of several people
> > who have been bitten by this very same problem.  It's worth noting
> > that the original author of the JMol message is a *tenured* faculty
> > member at U. Wisconsin, so not even tenure seems to be a complete
> > protection against the perception from many academic administrators
> > that contributing to open source scientific projects is "not scholarly
> > work".
>
> Yes, it is a very important problem. Indeed it would be nice to have a
> way of get scholar-like credit for work on open source projects. If this
> means creating a journal dedicated to released scientific code (let us
> just avoid putting "open source" in the title, it's too political), than
> it may be the way to go. I think it should not be restricted to Python
> and friends, a broader could increase its visibility.
>
> Any ideas welcome, there is a Damocles sword hanging above many
> contributions.
>
> Ga?l
> _______________________________________________
> 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/20071224/da556ea8/attachment.html>

From wnbell at gmail.com  Wed Dec 26 00:42:21 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Wed, 26 Dec 2007 05:42:21 +0000 (UTC)
Subject: [SciPy-user] arbitrary submatrices of a sparse matrix?
References: <C9311940-A4B9-4183-9F7C-1647DF7576C3@cs.toronto.edu>
	<9406141D-E419-4B6C-A4B4-033ACC2E1DF2@cs.toronto.edu>
Message-ID: <loom.20071226T051444-759@post.gmane.org>

David Warde-Farley <dwf <at> cs.toronto.edu> writes:

> > I've figured out how to do this for full matrices (either x
> > [[4,6,2,7],:][:,1,7,3]  or by playing with take()) but I'm at a loss
> > for how to do it with a sparse matrix, if it's even possible (not
> > sure how matlab pulls it off).
> 
> If there's some easy way to permute the rows and columns of a CSR/CSC  
> matrix then I would be able to get the job done with slices... Is  
> there? Or is there some way to do this that I'm not seeing?

You can use sparse matrix multiplication can to apply 
permutations and slice rows:

In [1]: from scipy import *
In [2]: from scipy.sparse import *
In [3]: A = csr_matrix(arange(12).reshape(4,3))
In [5]: A.todense()
Out[5]: 
matrix([[ 0,  1,  2],
        [ 3,  4,  5],
        [ 6,  7,  8],
        [ 9, 10, 11]])
In [7]: P = coo_matrix( ([1,1,1,1],[[0,1,2,3],[3,1,0,2]]) )
In [9]: (P*A).todense()
Out[9]: 
matrix([[ 9, 10, 11],
        [ 3,  4,  5],
        [ 0,  1,  2],
        [ 6,  7,  8]])
In [10]: P = coo_matrix( ([1,1,1,1],[[0,1,2,3],[0,0,3,3]]) )
In [11]: (P*A).todense()
Out[11]: 
matrix([[ 0,  1,  2],
        [ 0,  1,  2],
        [ 9, 10, 11],
        [ 9, 10, 11]])

The only downside of this approach is that it requires at least O(M) 
operations for an M,N sparse matrix in CSR format.  For permutations 
this is not a problem, and I believe you'll get near-optimal performance.
 However, if you only want to slice a few rows, then a special purpose 
approach is needed

I'm working on extending the work Robert has done to the sort of indexing 
you describe.  Keep in mind that the underlying matrix format limits the 
efficiency of slicing.  Slicing rows of a CSR or columns of a CSC is fast.
Slicing columns of a CSR is an O( A.nnz ) operation, which is essentially 
as bad as converting to CSC and then slicing.  I believe MATLAB uses CSC to
 represent sparse matrices, so one would expect A(:,10) to be faster than
A(10,:).  Here's a simple MATLAB example:

>> A = gallery('poisson',500);
>> tic; S = A(:,10); toc;
Elapsed time is 0.000060 seconds.
>> tic; S = A(10,:); toc;
Elapsed time is 0.063538 seconds.



From dmitrey.kroshko at scipy.org  Wed Dec 26 12:29:37 2007
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Wed, 26 Dec 2007 19:29:37 +0200
Subject: [SciPy-user] Would anyone connect fortran constrained linear least
 squares solver to Python?
In-Reply-To: <ccb1b5850712241447m62af6aeeh78f7b111f57e108e@mail.gmail.com>
References: <ccb1b5850712111941k41421e4ehdfedb1181a685879@mail.gmail.com>	<475F9EF7.4010509@scipy.org>	<ccb1b5850712161929n7d04d035l8676c95cf0895edd@mail.gmail.com>	<4768311C.7030906@scipy.org>	<ccb1b5850712201815u23602578sd4e142b4ab9069dc@mail.gmail.com>	<476B7C19.1020003@scipy.org>	<ccb1b5850712210604m7981ca68ufd6dd588c5fef93c@mail.gmail.com>	<ccb1b5850712210731p2cfdad86x6e782e1ed7b15a26@mail.gmail.com>	<476FE3F4.3000802@scipy.org>
	<ccb1b5850712241447m62af6aeeh78f7b111f57e108e@mail.gmail.com>
Message-ID: <47728F81.10906@scipy.org>

Hi all,

I had noticed (from traffic statistics): lots of people are interested 
in linear least squares problems (LLSP). However, scipy has only 
unconstrained LAPACK dGELSS/sGELSS.

Could anyone provide connection of the fortran-written solver to Python 
(or connect it to scipy)?

http://netlib3.cs.utk.edu/toms/587
(that one can handle linear eq and ineq constraints)

Then I would gladly provide connection of the one to scikits.openopt.

Regards, D.
http://scipy.org/scipy/scikits/wiki/Dmitrey



From lorenzo.isella at gmail.com  Wed Dec 26 19:28:12 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Thu, 27 Dec 2007 01:28:12 +0100
Subject: [SciPy-user] Scipy for Cluster Detection
Message-ID: <4772F19C.6040308@gmail.com>

Dear All,
I hope this will not sound too off-topic.
Without going into details, I am simulating the behavior of a set of 
particles interacting with a short-ranged, strongly binding potential. I 
start with a number N of particles which undergo collisions with each 
other and, in doing so, they stick giving rise to clusters.
The code which does that is not written in Python, but it returns the 
positions of these particles in space.
As time goes on, more and more particles coagulate around one of these 
clusters. Eventually, they will all end up in the same cluster.
Two particles are considered to be bounded if their distance falls below 
a certain threshold d.
A cluster is nothing else than a group of particles all directly or 
indirectly bound together.
Having said so, I now need an efficient algorithm to look for clusters, 
since at the very least I need to be able to count them to study how the 
number of clusters evolve in time.
Is there anything suitable for this already implemented in SciPy? I 
wonder if people in Astronomy have similar problems if they need e.g. to 
detect/study clusters of stars.
Simply I would like not to re-invent the wheel and it goes without 
saying that any suggestions here are really appreciated.
Many thanks and a nice Xmas break to everybody on this list.

Lorenzo



From william.ratcliff at gmail.com  Wed Dec 26 22:04:02 2007
From: william.ratcliff at gmail.com (william ratcliff)
Date: Wed, 26 Dec 2007 22:04:02 -0500
Subject: [SciPy-user] Scipy for Cluster Detection
In-Reply-To: <4772F19C.6040308@gmail.com>
References: <4772F19C.6040308@gmail.com>
Message-ID: <827183970712261904l4537914axe54657b7988bf717@mail.gmail.com>

Have you tried looking into the hierarchical clustering algorithms that are
in the biopython package?

http://www.inb.mu-luebeck.de/biosoft/biopython/api/Bio/Tools/Clustering/kMeans.py.html

Cheers,
William

On Dec 26, 2007 7:28 PM, Lorenzo Isella <lorenzo.isella at gmail.com> wrote:

> Dear All,
> I hope this will not sound too off-topic.
> Without going into details, I am simulating the behavior of a set of
> particles interacting with a short-ranged, strongly binding potential. I
> start with a number N of particles which undergo collisions with each
> other and, in doing so, they stick giving rise to clusters.
> The code which does that is not written in Python, but it returns the
> positions of these particles in space.
> As time goes on, more and more particles coagulate around one of these
> clusters. Eventually, they will all end up in the same cluster.
> Two particles are considered to be bounded if their distance falls below
> a certain threshold d.
> A cluster is nothing else than a group of particles all directly or
> indirectly bound together.
> Having said so, I now need an efficient algorithm to look for clusters,
> since at the very least I need to be able to count them to study how the
> number of clusters evolve in time.
> Is there anything suitable for this already implemented in SciPy? I
> wonder if people in Astronomy have similar problems if they need e.g. to
> detect/study clusters of stars.
> Simply I would like not to re-invent the wheel and it goes without
> saying that any suggestions here are really appreciated.
> Many thanks and a nice Xmas break to everybody on this list.
>
> Lorenzo
>
> _______________________________________________
> 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/20071226/7d6a93b2/attachment.html>

From dwf at cs.toronto.edu  Thu Dec 27 01:24:26 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 27 Dec 2007 01:24:26 -0500
Subject: [SciPy-user] Scipy for Cluster Detection
In-Reply-To: <4772F19C.6040308@gmail.com>
References: <4772F19C.6040308@gmail.com>
Message-ID: <C8CF9489-1457-46AF-B453-62A338B422B5@cs.toronto.edu>

On 26-Dec-07, at 7:28 PM, Lorenzo Isella wrote:

> Dear All,
> I hope this will not sound too off-topic.
> Without going into details, I am simulating the behavior of a set of
> particles interacting with a short-ranged, strongly binding  
> potential. I
> start with a number N of particles which undergo collisions with each
> other and, in doing so, they stick giving rise to clusters.
> The code which does that is not written in Python, but it returns the
> positions of these particles in space.
> As time goes on, more and more particles coagulate around one of these
> clusters. Eventually, they will all end up in the same cluster.
> Two particles are considered to be bounded if their distance falls  
> below
> a certain threshold d.
> A cluster is nothing else than a group of particles all directly or
> indirectly bound together.


This doesn't sound like what would typically be considered a

Do you have code that produces the (N^2 - N)/2 distances between pairs  
of particles? How large is N? If it's small enough, turning it into a  
NumPy array should be straightforward, and then thresholding at d is  
one line of code. Then what you end up with is a boolean matrix that  
where the (i,j) entry denotes that i and j are in a cluster together.  
This can be thought of as a graph or network where the nodes are  
particles and connections / links / edges exist between those  
particles that are in a cluster together.

Once you have this your problem is what graph theorists would call  
identifying the "connected components", which is a well-studied  
problem and can be done fairly easily, Google "finding connected  
components" or pick up any algorithms textbook at your local library  
such as the Cormen et al "Introduction to Algorithms".

I'm not in a position to comment on whether there's something like  
this already in SciPy, as I really don't know. However, finding  
connected components of a graph (i.e. your clusters) is a very well  
studied problem in the computer science (namely graph theory)  
literature, and the algorithms are simple enough that less than 20  
lines of Python should do the trick provided the distance matrix is  
manageable.

Regards,

David


From wnbell at gmail.com  Thu Dec 27 01:30:01 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Thu, 27 Dec 2007 00:30:01 -0600
Subject: [SciPy-user] Scipy for Cluster Detection
In-Reply-To: <4772F19C.6040308@gmail.com>
References: <4772F19C.6040308@gmail.com>
Message-ID: <d05265cb0712262230y3e94f982y9a0731a38bed4da0@mail.gmail.com>

On Dec 26, 2007 6:28 PM, Lorenzo Isella <lorenzo.isella at gmail.com> wrote:
> Two particles are considered to be bounded if their distance falls below
> a certain threshold d.

Assuming d is the same for all pairs of contacts you should probably
use spatial hashing.  I had a similar problem in the simulation of
granular materials and found spatial hashing to be the fastest
approach.
By spatial hashing I mean that to each point (x,y,z) you associate the
integer tuple ( floor( x/d ), (floor( y/d ), (floor( z/d ) ).  When
looking for all the neighbors of a point you simply look at all 27
neighboring tuples.  You could do this with Python dictionaries, but
you'll probably prefer something like the STL hash_map or Google's
sparsehash library.

If d varies then a k-D tree is probably your best best.

> A cluster is nothing else than a group of particles all directly or
> indirectly bound together.
> Having said so, I now need an efficient algorithm to look for clusters,
> since at the very least I need to be able to count them to study how the
> number of clusters evolve in time.
> Is there anything suitable for this already implemented in SciPy? I
> wonder if people in Astronomy have similar problems if they need e.g. to
> detect/study clusters of stars.
> Simply I would like not to re-invent the wheel and it goes without
> saying that any suggestions here are really appreciated.
> Many thanks and a nice Xmas break to everybody on this list.

I don't think SciPy has the functionality you need.  The cluster
module has k-means stuff, but I don't think that will help you.


-- 
Nathan Bell wnbell at gmail.com


From dwf at cs.toronto.edu  Thu Dec 27 01:36:53 2007
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 27 Dec 2007 01:36:53 -0500
Subject: [SciPy-user] Scipy for Cluster Detection
In-Reply-To: <C8CF9489-1457-46AF-B453-62A338B422B5@cs.toronto.edu>
References: <4772F19C.6040308@gmail.com>
	<C8CF9489-1457-46AF-B453-62A338B422B5@cs.toronto.edu>
Message-ID: <E493FE22-BED0-40C8-94BE-908B125AA043@cs.toronto.edu>


On 27-Dec-07, at 1:24 AM, David Warde-Farley wrote:
>
> This doesn't sound like what would typically be considered a

This should read, "This doesn't sound like what would typically be  
considered a clustering problem." What I meant is, "clustering"  
usually means finding a natural way of grouping the data (either  
hierarchically or non-hierarchically) where you only have some notion  
of similarity or distance to work with.
]
In your problem, your similarity metric is essentially binary ("are  
you less than d away from each other", yes or no) and so it reduces to  
a rather graph-theoretic problem.

That said, if your number of particles is particularly large, Nathan's  
suggestion of a kd-tree solution sounds like it's on the right track.

"Locality sensitive hashing" comes to mind, which is basically a fast  
way of finding the nearest neighbour of a given point -- if you knew  
this then you'd be able to tell whether any point at all was within d,  
since if the nearest neighbour isn't then none of them are.

Take a look at http://web.mit.edu/andoni/www/LSH/index.html

Cheers,

David


From matthieu.brucher at gmail.com  Thu Dec 27 01:42:01 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 27 Dec 2007 07:42:01 +0100
Subject: [SciPy-user] Scipy for Cluster Detection
In-Reply-To: <4772F19C.6040308@gmail.com>
References: <4772F19C.6040308@gmail.com>
Message-ID: <e76aa17f0712262242p7114b792y29b34393e6d93bbc@mail.gmail.com>

Hi,

Sounds like a pseudo correlation clustering problem, isn't ? It's not too
hard to create the clusters once you have the similarity matrix. You can
make a loop and for every point not labeled, you cluster around it every
point that are similar. This way you get a good approximation of your
optimal correlation problem (don't remember which article demonstrates it).

Matthieu

2007/12/27, Lorenzo Isella <lorenzo.isella at gmail.com>:
>
> Dear All,
> I hope this will not sound too off-topic.
> Without going into details, I am simulating the behavior of a set of
> particles interacting with a short-ranged, strongly binding potential. I
> start with a number N of particles which undergo collisions with each
> other and, in doing so, they stick giving rise to clusters.
> The code which does that is not written in Python, but it returns the
> positions of these particles in space.
> As time goes on, more and more particles coagulate around one of these
> clusters. Eventually, they will all end up in the same cluster.
> Two particles are considered to be bounded if their distance falls below
> a certain threshold d.
> A cluster is nothing else than a group of particles all directly or
> indirectly bound together.
> Having said so, I now need an efficient algorithm to look for clusters,
> since at the very least I need to be able to count them to study how the
> number of clusters evolve in time.
> Is there anything suitable for this already implemented in SciPy? I
> wonder if people in Astronomy have similar problems if they need e.g. to
> detect/study clusters of stars.
> Simply I would like not to re-invent the wheel and it goes without
> saying that any suggestions here are really appreciated.
> Many thanks and a nice Xmas break to everybody on this list.
>
> Lorenzo
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Website : http://matthieu-brucher.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/20071227/4f991563/attachment.html>

From sgarcia at olfac.univ-lyon1.fr  Thu Dec 27 03:54:57 2007
From: sgarcia at olfac.univ-lyon1.fr (Samuel GARCIA)
Date: Thu, 27 Dec 2007 09:54:57 +0100
Subject: [SciPy-user] Looking for super-paramagnetic clustering code.
Message-ID: <47736861.9020907@olfac.univ-lyon1.fr>

Hi list,
I am looking for a python implementation of super-paramagnetic 
clustering (SCP).

Here a link to the article : 
http://www.weizmann.ac.il/home/fedomany/papers.html

Does someones known about that ?
Is there other name for equivalent methods ?

Thanks

Samuel

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Samuel Garcia
Laboratoire de Neurosciences Sensorielles, Comportement, Cognition.
CNRS - UMR5020 - Universite Claude Bernard LYON 1
Equipe logistique et technique
50, avenue Tony Garnier
69366 LYON Cedex 07
FRANCE
T?l : 04 37 28 74 64
Fax : 04 37 28 76 01
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~



From vvinuv at gmail.com  Thu Dec 27 10:11:34 2007
From: vvinuv at gmail.com (Vinu Vikram)
Date: Thu, 27 Dec 2007 20:41:34 +0530
Subject: [SciPy-user] installation problem
Message-ID: <637563b70712270711j1853a42fm554afd7f805e035d@mail.gmail.com>

Hi all
I am facing problem when trying to install scipy0.6.0. The error I am
getting is
building 'scipy.integrate._odepack' extension
compiling C sources
C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
-Wstrict-prototypes -fPIC

compile options: '-DATLAS_INFO="\"3.6.0\""
-I/usr/local/lib/python2.5/site-packages/numpy/core/include
-I/usr/local/include/python2.5 -c'
/usr/bin/g77 -g -Wall -shared
build/temp.linux-x86_64-2.5/scipy/integrate/_odepackmodule.o
-L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2
-Lbuild/temp.linux-x86_64-2.5-lodepack -llinpack_lite -lmach
-lptf77blas -lptcblas -latlas -lg2c -o
build/lib.linux-x86_64-2.5/scipy/integrate/_odepack.so
/usr/bin/ld:
/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o):
relocation R_X86_64_32 against `a local symbol' can not be used when making
a shared object; recompile with -fPIC
/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not read
symbols: Bad value
collect2: ld returned 1 exit status
/usr/bin/ld:
/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o):
relocation R_X86_64_32 against `a local symbol' can not be used when making
a shared object; recompile with -fPIC
/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not read
symbols: Bad value
collect2: ld returned 1 exit status
error: Command "/usr/bin/g77 -g -Wall -shared
build/temp.linux-x86_64-2.5/scipy/integrate/_odepackmodule.o
-L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2
-Lbuild/temp.linux-x86_64-2.5-lodepack -llinpack_lite -lmach
-lptf77blas -lptcblas -latlas -lg2c -o
build/lib.linux-x86_64-2.5/scipy/integrate/_odepack.so" failed with exit
status 1

Please help me to solve this.
thanks
Vinu V

-- 
VINU VIKRAM
http://iucaa.ernet.in/~vvinuv/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071227/905a790a/attachment.html>

From cohen at slac.stanford.edu  Thu Dec 27 10:16:16 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Thu, 27 Dec 2007 07:16:16 -0800
Subject: [SciPy-user] installation problem
In-Reply-To: <637563b70712270711j1853a42fm554afd7f805e035d@mail.gmail.com>
References: <637563b70712270711j1853a42fm554afd7f805e035d@mail.gmail.com>
Message-ID: <4773C1C0.1060601@slac.stanford.edu>

hi Vinu,
you seem to have an installation of ATLAS which has not been built with 
the -Fa alg -fPIC flags (forcing position independent code for each 
compiler). See 
http://math-atlas.sourceforge.net/atlas_install/atlas_install.html#SECTION00043000000000000000

Johann

Vinu Vikram wrote:
> Hi all
> I am facing problem when trying to install scipy0.6.0. The error I am 
> getting is
> building 'scipy.integrate._odepack' extension
> compiling C sources
> C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall 
> -Wstrict-prototypes -fPIC
>
> compile options: '-DATLAS_INFO="\"3.6.0\"" 
> -I/usr/local/lib/python2.5/site-packages/numpy/core/include 
> -I/usr/local/include/python2.5 -c'
> /usr/bin/g77 -g -Wall -shared build/temp.linux-x86_64- 
> 2.5/scipy/integrate/_odepackmodule.o 
> -L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2 
> -Lbuild/temp.linux-x86_64-2.5 -lodepack -llinpack_lite -lmach 
> -lptf77blas -lptcblas -latlas -lg2c -o 
> build/lib.linux-x86_64-2.5/scipy/integrate/_odepack.so
> /usr/bin/ld: 
> /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o): 
> relocation R_X86_64_32 against `a local symbol' can not be used when 
> making a shared object; recompile with -fPIC
> /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not 
> read symbols: Bad value
> collect2: ld returned 1 exit status
> /usr/bin/ld: 
> /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o): 
> relocation R_X86_64_32 against `a local symbol' can not be used when 
> making a shared object; recompile with -fPIC
> /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not 
> read symbols: Bad value
> collect2: ld returned 1 exit status
> error: Command "/usr/bin/g77 -g -Wall -shared 
> build/temp.linux-x86_64-2.5/scipy/integrate/_odepackmodule.o 
> -L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2 -Lbuild/temp.linux-x86_64- 
> 2.5 -lodepack -llinpack_lite -lmach -lptf77blas -lptcblas -latlas 
> -lg2c -o build/lib.linux-x86_64-2.5/scipy/integrate/_odepack.so" 
> failed with exit status 1
>
> Please help me to solve this.
> thanks
> Vinu V
>
> -- 
> VINU VIKRAM
> http://iucaa.ernet.in/~vvinuv/ <http://iucaa.ernet.in/%7Evvinuv/>
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   


From lorenzo.isella at gmail.com  Thu Dec 27 11:44:45 2007
From: lorenzo.isella at gmail.com (Lorenzo Isella)
Date: Thu, 27 Dec 2007 17:44:45 +0100
Subject: [SciPy-user] Scipy for Cluster Detection
In-Reply-To: <mailman.5177.1198768299.13056.scipy-user@scipy.org>
References: <mailman.5177.1198768299.13056.scipy-user@scipy.org>
Message-ID: <4773D67D.8010701@gmail.com>

Hello,
Many thanks for the useful answers.
I should have added that I have in mind to deal with system with at most 
a few thousands of particles, each of them labeled by 3 coordinates.
The particle positions are read from a text file, then I can get the 
relative distances e.g. with some Fortran routine I import with f2py if 
speed is a problem.
If this is manageable, then I could really end up calculating directly 
the relative distances between all the particles and then do as 
suggested in the following.
I may post again early in January, however (I do not have enough time to 
test all of this straight away).
Cheers

Lorenzo
> Message: 3
> Date: Thu, 27 Dec 2007 01:24:26 -0500
> From: David Warde-Farley <dwf at cs.toronto.edu>
> Subject: Re: [SciPy-user] Scipy for Cluster Detection
> To: SciPy Users List <scipy-user at scipy.org>
> Message-ID: <C8CF9489-1457-46AF-B453-62A338B422B5 at cs.toronto.edu>
> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes
>
> On 26-Dec-07, at 7:28 PM, Lorenzo Isella wrote:
>
>   
>> Dear All,
>> I hope this will not sound too off-topic.
>> Without going into details, I am simulating the behavior of a set of
>> particles interacting with a short-ranged, strongly binding  
>> potential. I
>> start with a number N of particles which undergo collisions with each
>> other and, in doing so, they stick giving rise to clusters.
>> The code which does that is not written in Python, but it returns the
>> positions of these particles in space.
>> As time goes on, more and more particles coagulate around one of these
>> clusters. Eventually, they will all end up in the same cluster.
>> Two particles are considered to be bounded if their distance falls  
>> below
>> a certain threshold d.
>> A cluster is nothing else than a group of particles all directly or
>> indirectly bound together.
>>     
>
>
> This doesn't sound like what would typically be considered a
>
> Do you have code that produces the (N^2 - N)/2 distances between pairs  
> of particles? How large is N? If it's small enough, turning it into a  
> NumPy array should be straightforward, and then thresholding at d is  
> one line of code. Then what you end up with is a boolean matrix that  
> where the (i,j) entry denotes that i and j are in a cluster together.  
> This can be thought of as a graph or network where the nodes are  
> particles and connections / links / edges exist between those  
> particles that are in a cluster together.
>
> Once you have this your problem is what graph theorists would call  
> identifying the "connected components", which is a well-studied  
> problem and can be done fairly easily, Google "finding connected  
> components" or pick up any algorithms textbook at your local library  
> such as the Cormen et al "Introduction to Algorithms".
>
> I'm not in a position to comment on whether there's something like  
> this already in SciPy, as I really don't know. However, finding  
> connected components of a graph (i.e. your clusters) is a very well  
> studied problem in the computer science (namely graph theory)  
> literature, and the algorithms are simple enough that less than 20  
> lines of Python should do the trick provided the distance matrix is  
> manageable.
>
> Regards,
>
> David
>
>
>   



From timmichelsen at gmx-topmail.de  Thu Dec 27 12:45:35 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Thu, 27 Dec 2007 18:45:35 +0100
Subject: [SciPy-user] assignment of hours of day in time series
Message-ID: <fl0obv$tr8$1@ger.gmane.org>

Hello,
how do I assign the hours of a day in time series?

I have hourly measurements where hour 1 represents the end of the period 
0:00-1:00, 2 the end of the period 1:00-2:00, ... , 24 the end of the 
period 23:00 to 24:00.

When I plot these hourly time series from February to November the curve 
is continued into December because of that matter. time series then 
assumes that the value for hour 0:00 of dec, 01 is 0 which then leads do 
a wrong plotting behaviour.

I want to achieve that hour 24 is accounted as the last measurement 
period of a day and not as the first measurement of the next day (like 
0:00).

Please help me out here.

Thanks in advance,
Timmie



From fperez.net at gmail.com  Thu Dec 27 16:04:34 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Thu, 27 Dec 2007 14:04:34 -0700
Subject: [SciPy-user] Scipy server down...
Message-ID: <db6b5ecc0712271304t2b873371mce1131fc3ea21c58@mail.gmail.com>

Any chance it might be brought back up? (I'm trying to commit the
weave cleanup work Min did at the sprint...)

Thanks,

f


From millman at berkeley.edu  Thu Dec 27 16:11:37 2007
From: millman at berkeley.edu (Jarrod Millman)
Date: Thu, 27 Dec 2007 13:11:37 -0800
Subject: [SciPy-user] Scipy server down...
In-Reply-To: <db6b5ecc0712271304t2b873371mce1131fc3ea21c58@mail.gmail.com>
References: <db6b5ecc0712271304t2b873371mce1131fc3ea21c58@mail.gmail.com>
Message-ID: <c7009a550712271311wcf1cbd6tc37319d213e50c29@mail.gmail.com>

I restarted httpd.  It is still slow, but seems to be responsive now.
Let me know if it still isn't working.


On Dec 27, 2007 1:04 PM, Fernando Perez <fperez.net at gmail.com> wrote:
> Any chance it might be brought back up? (I'm trying to commit the
> weave cleanup work Min did at the sprint...)
>
> Thanks,
>
> f
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



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


From jeremit0 at gmail.com  Thu Dec 27 16:20:17 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Thu, 27 Dec 2007 16:20:17 -0500
Subject: [SciPy-user] Need very simple example showing how to return numpy
	array from C/C++ function
Message-ID: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>

I have a class member function that looks like this:
std::vector<double> discretize(...);

I would like to return a numpy array instead of a std::vector<double>.  I
could (of course) rewrite my function to return a pointer to the first
element of an array if needed.  I am currently using SWIG for wrapping my
code for Python.

I have searched far and wide for a simple example of how I should do this.
 Everything I find is *far* too complicated for someone doing this for the
first time.  Does someone have a simple example they would like to share?  I
have the "Guide to Numpy"; it has lots of information about the different
functions, it doesn't have any examples.

Thanks,
Jeremy Conlin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071227/7a77919e/attachment.html>

From fperez.net at gmail.com  Thu Dec 27 16:46:54 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Thu, 27 Dec 2007 14:46:54 -0700
Subject: [SciPy-user] Scipy server down...
In-Reply-To: <c7009a550712271311wcf1cbd6tc37319d213e50c29@mail.gmail.com>
References: <db6b5ecc0712271304t2b873371mce1131fc3ea21c58@mail.gmail.com>
	<c7009a550712271311wcf1cbd6tc37319d213e50c29@mail.gmail.com>
Message-ID: <db6b5ecc0712271346h2eaa3f57ocecffa88f6277f3a@mail.gmail.com>

On Dec 27, 2007 2:11 PM, Jarrod Millman <millman at berkeley.edu> wrote:
> I restarted httpd.  It is still slow, but seems to be responsive now.
> Let me know if it still isn't working.

SVN is working again, thanks.

Cheers,

f


From matthieu.brucher at gmail.com  Thu Dec 27 17:04:11 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Thu, 27 Dec 2007 23:04:11 +0100
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
Message-ID: <e76aa17f0712271404n7947eba9je632b721b51f59a8@mail.gmail.com>

Hi,

The problem is that you will have to copy yout data anyway. One of the best
solutions is to pass an already created Numpy array (as argument or with a
callback function) that you will be able to use (this is what I do with a
custom C++ array class that wraps the Python object).

Matthieu

2007/12/27, Jeremy Conlin <jeremit0 at gmail.com>:
>
> I have a class member function that looks like this:
> std::vector<double> discretize(...);
>
> I would like to return a numpy array instead of a std::vector<double>.  I
> could (of course) rewrite my function to return a pointer to the first
> element of an array if needed.  I am currently using SWIG for wrapping my
> code for Python.
>
> I have searched far and wide for a simple example of how I should do this.
>  Everything I find is *far* too complicated for someone doing this for the
> first time.  Does someone have a simple example they would like to share?  I
> have the "Guide to Numpy"; it has lots of information about the different
> functions, it doesn't have any examples.
>
> Thanks,
> Jeremy Conlin
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


-- 
French PhD student
Website : http://matthieu-brucher.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/20071227/a412d545/attachment.html>

From wnbell at gmail.com  Thu Dec 27 19:16:04 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Thu, 27 Dec 2007 18:16:04 -0600
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
Message-ID: <d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>

On Dec 27, 2007 3:20 PM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> I have a class member function that looks like this:
>
> std::vector<double> discretize(...);
>
> I would like to return a numpy array instead of a std::vector<double>.  I
> could (of course) rewrite my function to return a pointer to the first
> element of an array if needed.  I am currently using SWIG for wrapping my
> code for Python.
>
> I have searched far and wide for a simple example of how I should do this.
> Everything I find is *far* too complicated for someone doing this for the
> first time.  Does someone have a simple example they would like to share?  I
> have the "Guide to Numpy"; it has lots of information about the different
> functions, it doesn't have any examples.

I use STL vectors for ouput in scipy.sparse.  See line 517 of:
http://projects.scipy.org/scipy/scipy/browser/trunk/scipy/sparse/sparsetools/numpy.i

For a STL vector<double> named myvec this should work:

int length = myvec.size();
PyObject *obj = PyArray_FromDims(1, &length, PyArray_DOUBLE);
memcpy(PyArray_DATA(obj),&(myvec[0]),sizeof(double)*length);

If you know the size of the output in advance, then you might follow
Matthieu's suggestion since a copy is not required.

-- 
Nathan Bell wnbell at gmail.com


From bhendrix at enthought.com  Thu Dec 27 20:23:49 2007
From: bhendrix at enthought.com (Bryce Hendrix)
Date: Thu, 27 Dec 2007 19:23:49 -0600
Subject: [SciPy-user] [Numpy-discussion] Scipy server down...
In-Reply-To: <db6b5ecc0712271304t2b873371mce1131fc3ea21c58@mail.gmail.com>
References: <db6b5ecc0712271304t2b873371mce1131fc3ea21c58@mail.gmail.com>
Message-ID: <47745025.8080909@enthought.com>

Looks like the web site & svn are up for me.

Bryce

Fernando Perez wrote:
> Any chance it might be brought back up? (I'm trying to commit the
> weave cleanup work Min did at the sprint...)
>
> Thanks,
>
> f
> _______________________________________________
> Numpy-discussion mailing list
> Numpy-discussion at scipy.org
> http://projects.scipy.org/mailman/listinfo/numpy-discussion
>
>   



From vvinuv at gmail.com  Thu Dec 27 23:14:34 2007
From: vvinuv at gmail.com (Vinu Vikram)
Date: Fri, 28 Dec 2007 09:44:34 +0530
Subject: [SciPy-user] installation problem
In-Reply-To: <4773C1C0.1060601@slac.stanford.edu>
References: <637563b70712270711j1853a42fm554afd7f805e035d@mail.gmail.com>
	<4773C1C0.1060601@slac.stanford.edu>
Message-ID: <637563b70712272014g4401f1a6nca1a399f12f0e8d9@mail.gmail.com>

I worked . Thanks!!
there was no error during the installation. But when I tried to import
scipy.signal, I got the following error

>>> import scipy.signal
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python2.5/site-packages/scipy/signal/__init__.py",
line 10, in <module>
    from filter_design import *
  File
"/usr/local/lib/python2.5/site-packages/scipy/signal/filter_design.py", line
9, in <module>
    from scipy import special, optimize
  File "/usr/local/lib/python2.5/site-packages/scipy/optimize/__init__.py",
line 11, in <module>
    from lbfgsb import fmin_l_bfgs_b
  File "/usr/local/lib/python2.5/site-packages/scipy/optimize/lbfgsb.py",
line 30, in <module>
    import _lbfgsb
ImportError:
/usr/local/lib/python2.5/site-packages/scipy/optimize/_lbfgsb.so: undefined
symbol: _gfortran_st_write_done

How I can fix it?
Thanks
Vinu V

On Dec 27, 2007 8:46 PM, Johann Cohen-Tanugi <cohen at slac.stanford.edu>
wrote:

> hi Vinu,
> you seem to have an installation of ATLAS which has not been built with
> the -Fa alg -fPIC flags (forcing position independent code for each
> compiler). See
>
> http://math-atlas.sourceforge.net/atlas_install/atlas_install.html#SECTION00043000000000000000
>
> Johann
>
> Vinu Vikram wrote:
> > Hi all
> > I am facing problem when trying to install scipy0.6.0. The error I am
> > getting is
> > building 'scipy.integrate._odepack' extension
> > compiling C sources
> > C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
> > -Wstrict-prototypes -fPIC
> >
> > compile options: '-DATLAS_INFO="\"3.6.0\""
> > -I/usr/local/lib/python2.5/site-packages/numpy/core/include
> > -I/usr/local/include/python2.5 -c'
> > /usr/bin/g77 -g -Wall -shared build/temp.linux-x86_64-
> > 2.5/scipy/integrate/_odepackmodule.o
> > -L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2
> > -Lbuild/temp.linux-x86_64-2.5 -lodepack -llinpack_lite -lmach
> > -lptf77blas -lptcblas -latlas -lg2c -o
> > build/lib.linux-x86_64-2.5/scipy/integrate/_odepack.so
> > /usr/bin/ld:
> >
> /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o):
> > relocation R_X86_64_32 against `a local symbol' can not be used when
> > making a shared object; recompile with -fPIC
> > /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not
> > read symbols: Bad value
> > collect2: ld returned 1 exit status
> > /usr/bin/ld:
> >
> /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o):
> > relocation R_X86_64_32 against `a local symbol' can not be used when
> > making a shared object; recompile with -fPIC
> > /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not
> > read symbols: Bad value
> > collect2: ld returned 1 exit status
> > error: Command "/usr/bin/g77 -g -Wall -shared
> > build/temp.linux-x86_64-2.5/scipy/integrate/_odepackmodule.o
> > -L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2 -Lbuild/temp.linux-x86_64-
> > 2.5 -lodepack -llinpack_lite -lmach -lptf77blas -lptcblas -latlas
> > -lg2c -o build/lib.linux-x86_64-2.5/scipy/integrate/_odepack.so"
> > failed with exit status 1
> >
> > Please help me to solve this.
> > thanks
> > Vinu V
> >
> > --
> > VINU VIKRAM
> > http://iucaa.ernet.in/~vvinuv/ <http://iucaa.ernet.in/%7Evvinuv/> <
> http://iucaa.ernet.in/%7Evvinuv/>
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > 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
>



-- 
VINU VIKRAM
http://iucaa.ernet.in/~vvinuv/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071228/f3038fc4/attachment.html>

From cohen at slac.stanford.edu  Fri Dec 28 00:06:05 2007
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Thu, 27 Dec 2007 21:06:05 -0800
Subject: [SciPy-user] installation problem
In-Reply-To: <637563b70712272014g4401f1a6nca1a399f12f0e8d9@mail.gmail.com>
References: <637563b70712270711j1853a42fm554afd7f805e035d@mail.gmail.com>	<4773C1C0.1060601@slac.stanford.edu>
	<637563b70712272014g4401f1a6nca1a399f12f0e8d9@mail.gmail.com>
Message-ID: <4774843D.3020005@slac.stanford.edu>

hmm I don't have this symbol in my version of this shared lib, and I am 
no expert here :) But did you make sure that you  compiled ATLAS and 
scipy with gfortran? I could imagine something like that happening if 
scipy was built with g77 and ATLAS with gfortran.... You can use the 
fcompiler option to setup.py to ensure that gfortran is picked up as the 
fortran compiler.
Maybe an expert will kick in here...
hth,
Johann

Vinu Vikram wrote:
> I worked . Thanks!!
> there was no error during the installation. But when I tried to import 
> scipy.signal, I got the following error
>
> >>> import scipy.signal
> Traceback (most recent call last):
>   File "<stdin>", line 1, in <module>
>   File 
> "/usr/local/lib/python2.5/site-packages/scipy/signal/__init__.py", 
> line 10, in <module>
>     from filter_design import *
>   File 
> "/usr/local/lib/python2.5/site-packages/scipy/signal/filter_design.py", 
> line 9, in <module>
>     from scipy import special, optimize
>   File 
> "/usr/local/lib/python2.5/site-packages/scipy/optimize/__init__.py", 
> line 11, in <module>
>     from lbfgsb import fmin_l_bfgs_b
>   File 
> "/usr/local/lib/python2.5/site-packages/scipy/optimize/lbfgsb.py", 
> line 30, in <module>
>     import _lbfgsb
> ImportError: 
> /usr/local/lib/python2.5/site-packages/scipy/optimize/_lbfgsb.so: 
> undefined symbol: _gfortran_st_write_done
>
> How I can fix it?
> Thanks
> Vinu V
>
> On Dec 27, 2007 8:46 PM, Johann Cohen-Tanugi <cohen at slac.stanford.edu 
> <mailto:cohen at slac.stanford.edu>> wrote:
>
>     hi Vinu,
>     you seem to have an installation of ATLAS which has not been built
>     with
>     the -Fa alg -fPIC flags (forcing position independent code for each
>     compiler). See
>     http://math-atlas.sourceforge.net/atlas_install/atlas_install.html#SECTION00043000000000000000
>
>     Johann
>
>     Vinu Vikram wrote:
>     > Hi all
>     > I am facing problem when trying to install scipy0.6.0. The error
>     I am
>     > getting is
>     > building 'scipy.integrate._odepack' extension
>     > compiling C sources
>     > C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
>     > -Wstrict-prototypes -fPIC
>     >
>     > compile options: '-DATLAS_INFO="\"3.6.0\""
>     > -I/usr/local/lib/python2.5/site-packages/numpy/core/include
>     > -I/usr/local/include/python2.5 -c'
>     > /usr/bin/g77 -g -Wall -shared build/temp.linux-x86_64-
>     > 2.5/scipy/integrate/_odepackmodule.o
>     > -L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2
>     > -Lbuild/temp.linux-x86_64-2.5 -lodepack -llinpack_lite -lmach
>     > -lptf77blas -lptcblas -latlas -lg2c -o
>     > build/lib.linux-x86_64- 2.5/scipy/integrate/_odepack.so
>     > /usr/bin/ld:
>     >
>     /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o):
>     > relocation R_X86_64_32 against `a local symbol' can not be used
>     when
>     > making a shared object; recompile with -fPIC
>     > /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not
>     > read symbols: Bad value
>     > collect2: ld returned 1 exit status
>     > /usr/bin/ld:
>     >
>     /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a(ATL_F77wrap_dptscal.o):
>     > relocation R_X86_64_32 against `a local symbol' can not be used when
>     > making a shared object; recompile with -fPIC
>     > /home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2/libptf77blas.a: could not
>     > read symbols: Bad value
>     > collect2: ld returned 1 exit status
>     > error: Command "/usr/bin/g77 -g -Wall -shared
>     > build/temp.linux-x86_64- 2.5/scipy/integrate/_odepackmodule.o
>     > -L/home/vinu/ATLAS/lib/Linux_UNKNOWNSSE2_2
>     -Lbuild/temp.linux-x86_64-
>     > 2.5 -lodepack -llinpack_lite -lmach -lptf77blas -lptcblas -latlas
>     > -lg2c -o build/lib.linux-x86_64- 2.5/scipy/integrate/_odepack.so"
>     > failed with exit status 1
>     >
>     > Please help me to solve this.
>     > thanks
>     > Vinu V
>     >
>     > --
>     > VINU VIKRAM
>     > http://iucaa.ernet.in/~vvinuv/
>     <http://iucaa.ernet.in/%7Evvinuv/> <http://iucaa.ernet.in/%7Evvinuv/>
>     >
>     ------------------------------------------------------------------------
>     >
>     > _______________________________________________
>     > SciPy-user mailing list
>     > SciPy-user at scipy.org <mailto:SciPy-user at scipy.org>
>     > http://projects.scipy.org/mailman/listinfo/scipy-user
>     >
>     _______________________________________________
>     SciPy-user mailing list
>     SciPy-user at scipy.org <mailto:SciPy-user at scipy.org>
>     http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>
> -- 
> VINU VIKRAM
> http://iucaa.ernet.in/~vvinuv/ <http://iucaa.ernet.in/%7Evvinuv/>
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   


From matthieu.brucher at gmail.com  Fri Dec 28 04:32:10 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 28 Dec 2007 10:32:10 +0100
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
	<d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
Message-ID: <e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>

2007/12/28, Nathan Bell <wnbell at gmail.com>:
>
> On Dec 27, 2007 3:20 PM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> > I have a class member function that looks like this:
> >
> > std::vector<double> discretize(...);
> >
> > I would like to return a numpy array instead of a
> std::vector<double>.  I
> > could (of course) rewrite my function to return a pointer to the first
> > element of an array if needed.  I am currently using SWIG for wrapping
> my
> > code for Python.
> >
> > I have searched far and wide for a simple example of how I should do
> this.
> > Everything I find is *far* too complicated for someone doing this for
> the
> > first time.  Does someone have a simple example they would like to
> share?  I
> > have the "Guide to Numpy"; it has lots of information about the
> different
> > functions, it doesn't have any examples.
>
> I use STL vectors for ouput in scipy.sparse.  See line 517 of:
>
> http://projects.scipy.org/scipy/scipy/browser/trunk/scipy/sparse/sparsetools/numpy.i
>
> For a STL vector<double> named myvec this should work:
>
> int length = myvec.size();
> PyObject *obj = PyArray_FromDims(1, &length, PyArray_DOUBLE);
> memcpy(PyArray_DATA(obj),&(myvec[0]),sizeof(double)*length);
>
> If you know the size of the output in advance, then you might follow
> Matthieu's suggestion since a copy is not required.
>

You can then create a custom typemap with Nathan's code (not tested) :
%typemap(out) std::vector<double> %{
int length = $1.size();
$result = PyArray_FromDims(1, &length, PyArray_DOUBLE);
memcpy(PyArray_DATA($result),&($1[0]),sizeof(double)*length);
%}

The $1 indicates the data that will be wrapped and copied and $result is the
corresponding Python object.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071228/de400a86/attachment.html>

From timmichelsen at gmx-topmail.de  Fri Dec 28 10:34:06 2007
From: timmichelsen at gmx-topmail.de (Tim Michelsen)
Date: Fri, 28 Dec 2007 16:34:06 +0100
Subject: [SciPy-user] end_date doesn't work in time series
Message-ID: <fl351e$nge$2@ger.gmane.org>

Hello!

According to the descriptions in the wiki 
(http://www.scipy.org/SciPyPackages/TimeSeries#head-352d5481446b38e165dff3993e39f49c0e902a91-2) 
one should be able to define a time series by giving start /and/ end date.

like:
mydata_ts_hourly = TS.time_series(mydata, start_date=D_hr_start, 
end_date=None)

But defining a time series this way doesn't work:
TypeError: time_series() got an unexpected keyword argument 'end_date'

How can I get this work?
Is it a bug and where shall I report this?

Kind regards,
Timmie



From jeremit0 at gmail.com  Fri Dec 28 13:58:23 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Fri, 28 Dec 2007 13:58:23 -0500
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
	<d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
	<e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>
Message-ID: <3db594f70712281058vb291fb1u1604ff8cbaaca204@mail.gmail.com>

On Dec 28, 2007 4:32 AM, Matthieu Brucher <matthieu.brucher at gmail.com>
wrote:

>
>
> 2007/12/28, Nathan Bell <wnbell at gmail.com>:
> >
> > On Dec 27, 2007 3:20 PM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
> > > I have a class member function that looks like this:
> > >
> > > std::vector<double> discretize(...);
> > >
> > > I would like to return a numpy array instead of a
> > std::vector<double>.  I
> > > could (of course) rewrite my function to return a pointer to the first
> > > element of an array if needed.  I am currently using SWIG for wrapping
> > my
> > > code for Python.
> > >
> > > I have searched far and wide for a simple example of how I should do
> > this.
> > > Everything I find is *far* too complicated for someone doing this for
> > the
> > > first time.  Does someone have a simple example they would like to
> > share?  I
> > > have the "Guide to Numpy"; it has lots of information about the
> > different
> > > functions, it doesn't have any examples.
> >
> > I use STL vectors for ouput in scipy.sparse.  See line 517 of:
> >
> > http://projects.scipy.org/scipy/scipy/browser/trunk/scipy/sparse/sparsetools/numpy.i
> >
> > For a STL vector<double> named myvec this should work:
> >
> > int length = myvec.size();
> > PyObject *obj = PyArray_FromDims(1, &length, PyArray_DOUBLE);
> > memcpy(PyArray_DATA(obj),&(myvec[0]),sizeof(double)*length);
> >
> > If you know the size of the output in advance, then you might follow
> > Matthieu's suggestion since a copy is not required.
> >
>
> You can then create a custom typemap with Nathan's code (not tested) :
> %typemap(out) std::vector<double> %{
> int length = $1.size();
> $result = PyArray_FromDims(1, &length, PyArray_DOUBLE);
> memcpy(PyArray_DATA($result),&($1[0]),sizeof(double)*length);
> %}
>


Thank you, that is beginning to make more sense to one inexperienced in
wrapping.  There is one error that I haven't been able to work around.

error: no match for 'operator[]' in 'result[0]'

This comes from the memcpy line, I guess.  It looks like it is getting mixed
up between the $1 and the $result.  Do I need to apply this typemap to avoid
this error or is it automatically applied because the type (double) is
specified?

Thanks again,
Jeremy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071228/b0e27a80/attachment.html>

From matthieu.brucher at gmail.com  Fri Dec 28 14:06:24 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 28 Dec 2007 20:06:24 +0100
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <3db594f70712281058vb291fb1u1604ff8cbaaca204@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
	<d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
	<e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>
	<3db594f70712281058vb291fb1u1604ff8cbaaca204@mail.gmail.com>
Message-ID: <e76aa17f0712281106u769b5012j4ff997b6b480c739@mail.gmail.com>

>
> Thank you, that is beginning to make more sense to one inexperienced in
> wrapping.  There is one error that I haven't been able to work around.
>
> error: no match for 'operator[]' in 'result[0]'
>
> This comes from the memcpy line, I guess.  It looks like it is getting
> mixed up between the $1 and the $result.
>


This is strange... Is there more indications like the type of result ?


Do I need to apply this typemap to avoid this error or is it automatically
> applied because the type (double) is specified?
>


The fact that you have a compile error shows that it is applied ;)

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071228/703266bc/attachment.html>

From pgmdevlist at gmail.com  Fri Dec 28 14:08:30 2007
From: pgmdevlist at gmail.com (Pierre GM)
Date: Fri, 28 Dec 2007 14:08:30 -0500
Subject: [SciPy-user] end_date doesn't work in time series
In-Reply-To: <fl351e$nge$2@ger.gmane.org>
References: <fl351e$nge$2@ger.gmane.org>
Message-ID: <200712281408.37524.pgmdevlist@gmail.com>

Tim,

> According to the descriptions in the wiki
> (http://www.scipy.org/SciPyPackages/TimeSeries#head-352d5481446b38e165dff39
>93e39f49c0e902a91-2) one should be able to define a time series by giving
> start /and/ end date.

Well, sorry about that: I need to update the doc. Matt and I decided to get 
rid of the end_date keyword, as it brings more problems than it solves. 
Basically, you only need to provide a starting date, the end date will be 
defined according to the size of the input data. So, your command:

> mydata_ts_hourly = TS.time_series(mydata, start_date=D_hr_start,
> end_date=None)

should in fact be 

mydata_ts_hourly = TS.time_series(mydata, start_date=D_hr_start)


From jeremit0 at gmail.com  Fri Dec 28 15:44:43 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Fri, 28 Dec 2007 15:44:43 -0500
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <e76aa17f0712281106u769b5012j4ff997b6b480c739@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
	<d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
	<e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>
	<3db594f70712281058vb291fb1u1604ff8cbaaca204@mail.gmail.com>
	<e76aa17f0712281106u769b5012j4ff997b6b480c739@mail.gmail.com>
Message-ID: <3db594f70712281244i45baf536y358cb28ee6e79f42@mail.gmail.com>

On Dec 28, 2007 2:06 PM, Matthieu Brucher <matthieu.brucher at gmail.com>
wrote:

> Thank you, that is beginning to make more sense to one inexperienced in
> > wrapping.  There is one error that I haven't been able to work around.
> >
> > error: no match for 'operator[]' in 'result[0]'
> >
> > This comes from the memcpy line, I guess.  It looks like it is getting
> > mixed up between the $1 and the $result.
> >
>
>
> This is strange... Is there more indications like the type of result ?
>
>
There wasn't much else in the compiler messages.  I am just figuring this
out before I add it to my real code.  As such, the files I use are simple.
 I have attached them to this email.


> Do I need to apply this typemap to avoid this error or is it automatically
> > applied because the type (double) is specified?
> >
>
>
> The fact that you have a compile error shows that it is applied ;)
>
Of course!

Thanks again,
Jeremy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071228/8c624efc/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_SWIG.i
Type: application/octet-stream
Size: 678 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071228/8c624efc/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_SWIG.h
Type: application/octet-stream
Size: 527 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071228/8c624efc/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_SWIG.cpp
Type: application/octet-stream
Size: 594 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071228/8c624efc/attachment-0002.obj>

From humufr at yahoo.fr  Fri Dec 28 15:46:32 2007
From: humufr at yahoo.fr (humufr at yahoo.fr)
Date: Fri, 28 Dec 2007 15:46:32 -0500
Subject: [SciPy-user] maskedarray moved?
Message-ID: <200712281546.33014.humufr@yahoo.fr>

		Hi,

I would like to know where is maskedarray. It's not anymore inside the 
sandbox? Where can I find it now?

Thanks,

N.


From millman at berkeley.edu  Fri Dec 28 16:03:37 2007
From: millman at berkeley.edu (Jarrod Millman)
Date: Fri, 28 Dec 2007 13:03:37 -0800
Subject: [SciPy-user] maskedarray moved?
In-Reply-To: <200712281546.33014.humufr@yahoo.fr>
References: <200712281546.33014.humufr@yahoo.fr>
Message-ID: <c7009a550712281303k5a134712kc4580e46bd224650@mail.gmail.com>

It is now in a numpy branch:
http://projects.scipy.org/scipy/numpy/browser/branches/maskedarray

and will be merged with the numpy trunk soon.  You can view the API
changes with the existing numpy.ma here:
http://svn.scipy.org/svn/numpy/branches/maskedarray/numpy/ma/API_CHANGES.txt

Thanks,
Jarrod

On Dec 28, 2007 12:46 PM,  <humufr at yahoo.fr> wrote:
>                 Hi,
>
> I would like to know where is maskedarray. It's not anymore inside the
> sandbox? Where can I find it now?
>
> Thanks,
>
> N.
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
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  Fri Dec 28 17:50:59 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 28 Dec 2007 23:50:59 +0100
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <3db594f70712281244i45baf536y358cb28ee6e79f42@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
	<d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
	<e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>
	<3db594f70712281058vb291fb1u1604ff8cbaaca204@mail.gmail.com>
	<e76aa17f0712281106u769b5012j4ff997b6b480c739@mail.gmail.com>
	<3db594f70712281244i45baf536y358cb28ee6e79f42@mail.gmail.com>
Message-ID: <e76aa17f0712281450r1c0a8a22ob0dd079d5844d8a2@mail.gmail.com>

To make your .i work, you have two things to do :
- first initialize numpy :
%{
#define SWIG_FILE_WITH_INIT
%}
%include "numpy.i"
%init %{
import_array();
%}

Then result (which is $1 in the typemap) is in fact a wrapper around the
vector, and to get it, you have to dereference it, so the typemap becomes :
%typemap(out) std::vector<double> {
    int length = $1.size();
    $result = PyArray_FromDims(1, &length, PyArray_DOUBLE);
    memcpy(PyArray_DATA($result),&((*(&$1))[0]),sizeof(double)*length);
}

I think I will add this one to my blog, if nobody minds.

Matthieu

2007/12/28, Jeremy Conlin <jeremit0 at gmail.com>:
>
>
>
> On Dec 28, 2007 2:06 PM, Matthieu Brucher <matthieu.brucher at gmail.com>
> wrote:
>
> > Thank you, that is beginning to make more sense to one inexperienced in
> > > wrapping.  There is one error that I haven't been able to work around.
> > >
> > > error: no match for 'operator[]' in 'result[0]'
> > >
> > > This comes from the memcpy line, I guess.  It looks like it is getting
> > > mixed up between the $1 and the $result.
> > >
> >
> >
> > This is strange... Is there more indications like the type of result ?
> >
> >
> There wasn't much else in the compiler messages.  I am just figuring this
> out before I add it to my real code.  As such, the files I use are simple.
>  I have attached them to this email.
>
>
> > Do I need to apply this typemap to avoid this error or is it
> > > automatically applied because the type (double) is specified?
> > >
> >
> >
> > The fact that you have a compile error shows that it is applied ;)
> >
> Of course!
>
> Thanks again,
> Jeremy
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>


-- 
French PhD student
Website : http://matthieu-brucher.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/20071228/e1830727/attachment.html>

From ellisonbg.net at gmail.com  Fri Dec 28 18:15:43 2007
From: ellisonbg.net at gmail.com (Brian Granger)
Date: Fri, 28 Dec 2007 16:15:43 -0700
Subject: [SciPy-user] Banded matrix wrapper
Message-ID: <6ce0ac130712281515k11f92714i2fb863e12704d1cc@mail.gmail.com>

Hi,

I am needing to solve an eigenvalue problem using banded matrices.  I
see that scipy now has a function for doing this, but it takes the
matrix in a particular format (the LAPACK banded matrix format).  Does
anyone have any utilities for working with this format?  I figured I
would ask before doing this myself.

Thanks

Brian


From jeremit0 at gmail.com  Fri Dec 28 19:07:59 2007
From: jeremit0 at gmail.com (Jeremy Conlin)
Date: Fri, 28 Dec 2007 19:07:59 -0500
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <e76aa17f0712281450r1c0a8a22ob0dd079d5844d8a2@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
	<d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
	<e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>
	<3db594f70712281058vb291fb1u1604ff8cbaaca204@mail.gmail.com>
	<e76aa17f0712281106u769b5012j4ff997b6b480c739@mail.gmail.com>
	<3db594f70712281244i45baf536y358cb28ee6e79f42@mail.gmail.com>
	<e76aa17f0712281450r1c0a8a22ob0dd079d5844d8a2@mail.gmail.com>
Message-ID: <3db594f70712281607n4f79d6a5w9d946c30f08dcc39@mail.gmail.com>

On Dec 28, 2007 5:50 PM, Matthieu Brucher <matthieu.brucher at gmail.com>
wrote:

> To make your .i work, you have two things to do :
> - first initialize numpy :
> %{
> #define SWIG_FILE_WITH_INIT
> %}
> %include "numpy.i"
> %init %{
> import_array();
> %}
>
> Then result (which is $1 in the typemap) is in fact a wrapper around the
> vector, and to get it, you have to dereference it, so the typemap becomes :
> %typemap(out) std::vector<double> {
>     int length = $1.size();
>     $result = PyArray_FromDims(1, &length, PyArray_DOUBLE);
>     memcpy(PyArray_DATA($result),&((*(&$1))[0]),sizeof(double)*length);
> }
>

YEAH!!! Now it works!  Thanks a lot!

Now that it works, I wonder how this could be done without copying the data.
 Since a STL vector can also be accessed and treated like an array, could we
exploit this to avoid copying?  I tried using the function
PyArray_FromDimsAndData and passing a reference to the first element using
the same syntax as for memcpy, but I got an error similar to the one I
previously reported:

error: no match for 'operator*' in '*result'



>
> I think I will add this one to my blog, if nobody minds.

I think it would be great to have a simple example like this available to
the community.  A page on the wiki would probably be best.  I could create
such a page, but with my limited experience, I'm not sure I would be the
best person.  I couldn't explain why all the elements in the *.i file are
necessary.

Thanks,
Jeremy


>
> Matthieu
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071228/09d82efe/attachment.html>

From bnuttall at uky.edu  Fri Dec 28 19:06:57 2007
From: bnuttall at uky.edu (Nuttall, Brandon C)
Date: Fri, 28 Dec 2007 19:06:57 -0500
Subject: [SciPy-user] assignment of hours of day in time series
In-Reply-To: <fl0obv$tr8$1@ger.gmane.org>
References: <fl0obv$tr8$1@ger.gmane.org>
Message-ID: <F5BC85B8534D524DB972D2B2731CA53F04CFB7CE3A@EX7FM01.ad.uky.edu>

Timmie,

I'd start with the Python built-in datetime module and see if that helps.

Brandon

________________________________________
From: scipy-user-bounces at scipy.org [scipy-user-bounces at scipy.org] On Behalf Of Tim Michelsen [timmichelsen at gmx-topmail.de]
Sent: Thursday, December 27, 2007 12:45 PM
To: scipy-user at scipy.org
Subject: [SciPy-user] assignment of hours of day in time series

Hello,
how do I assign the hours of a day in time series?

I have hourly measurements where hour 1 represents the end of the period
0:00-1:00, 2 the end of the period 1:00-2:00, ... , 24 the end of the
period 23:00 to 24:00.

When I plot these hourly time series from February to November the curve
is continued into December because of that matter. time series then
assumes that the value for hour 0:00 of dec, 01 is 0 which then leads do
a wrong plotting behaviour.

I want to achieve that hour 24 is accounted as the last measurement
period of a day and not as the first measurement of the next day (like
0:00).

Please help me out here.

Thanks in advance,
Timmie

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


From wnbell at gmail.com  Fri Dec 28 19:57:45 2007
From: wnbell at gmail.com (Nathan Bell)
Date: Fri, 28 Dec 2007 18:57:45 -0600
Subject: [SciPy-user] Need very simple example showing how to return
	numpy array from C/C++ function
In-Reply-To: <3db594f70712281607n4f79d6a5w9d946c30f08dcc39@mail.gmail.com>
References: <3db594f70712271320l18ed1c3fh728ff05c8ceae66d@mail.gmail.com>
	<d05265cb0712271616i433f8b5fr4913e582b29767a0@mail.gmail.com>
	<e76aa17f0712280132p5866e97dh15950ebf54b165cc@mail.gmail.com>
	<3db594f70712281058vb291fb1u1604ff8cbaaca204@mail.gmail.com>
	<e76aa17f0712281106u769b5012j4ff997b6b480c739@mail.gmail.com>
	<3db594f70712281244i45baf536y358cb28ee6e79f42@mail.gmail.com>
	<e76aa17f0712281450r1c0a8a22ob0dd079d5844d8a2@mail.gmail.com>
	<3db594f70712281607n4f79d6a5w9d946c30f08dcc39@mail.gmail.com>
Message-ID: <d05265cb0712281657r65392eb4g422f4219e43dd913@mail.gmail.com>

On Dec 28, 2007 6:07 PM, Jeremy Conlin <jeremit0 at gmail.com> wrote:
>
> Now that it works, I wonder how this could be done without copying the data.
> Since a STL vector can also be accessed and treated like an array, could we
> exploit this to avoid copying?  I tried using the function
> PyArray_FromDimsAndData and passing a reference to the first element using
> the same syntax as for memcpy, but I got an error similar to the one I
> previously reported:

Ideally you could pluck out the contiguous memory chunk used by the
STL vector and hand that to numpy.  However I doubt this would work
because the vector probably needs to free its data to a different
allocator (I would assume numpy uses malloc and the STL uses some C++
weirdness).

> I think it would be great to have a simple example like this available to
> the community.  A page on the wiki would probably be best.  I could create
> such a page, but with my limited experience, I'm not sure I would be the
> best person.  I couldn't explain why all the elements in the *.i file are
> necessary.

Agreed.  I'd like to see if anyone else comes up with better ways of
interfacing C++ and scipy.

Matthieu, if you have time you might also include a pointer to these
wrappers for numpy's complex types:
http://projects.scipy.org/scipy/scipy/browser/trunk/scipy/sparse/sparsetools/complex_ops.h



-- 
Nathan Bell wnbell at gmail.com


From mani.sabri at gmail.com  Sat Dec 29 03:28:51 2007
From: mani.sabri at gmail.com (mani sabri)
Date: Sat, 29 Dec 2007 11:58:51 +0330
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
Message-ID: <4776059b.03be100a.54cf.2160@mx.google.com>

Hi 

Happy holidays

I'm using python 2.4.4 , scipy 0.6.0 , numpy 1.0.4  and mingw gcc 3.4.5 on
win xp sp2 

When I try to run a weave example (e.x array3d.py) from "C:\PROGRAM
FILES\Python24\Lib\site-packages\scipy\weave\examples" I have the following
typical output:

 

numpy:

[[[ 0  1  2  3]

  [ 4  5  6  7]

  [ 8  9 10 11]]

 

 [[12 13 14 15]

  [16 17 18 19]

  [20 21 22 23]]]

 

Pure Inline:

<weave: compiling>

Found executable C:\MINGW\BIN\g++.exe

g++.exe: C:\program: No such file or directory

g++.exe: files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp: No
such file or directory

g++.exe: files\python24\lib\site-packages\scipy\weave\scxx\weave_imp.o: No
such file or directory

g++.exe: no input files

 

And This trace back:

 

CompileError: error: Command "g++ -mno-cygwin -O2 -Wall -I"C:\program
files\Python24\lib\site-packages\scipy\weave" -I"C:\program
files\Python24\lib\site-packages\scipy\weave\scxx" -I"C:\program
files\Python24\lib\site-packages\numpy\core\include" -I"C:\program
files\Python24\include" -I"C:\program files\Python24\PC" -c C:\program
files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp -o
c:\docume~1\man\locals~1\temp\man\python24_intermediate\compiler_12e837eb1ea
3ab5199fbcc0e83015e3f\Release\program
files\python24\lib\site-packages\scipy\weave\scxx\weave_imp.o" failed with
exit status 1

 

IMHO When I put quotes around -c argument  in above command (-c "C:\program
files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp") every
thing is fine.

Can anyone help me to hack this around?

 

Cheers

Mani.

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071229/f3d1c46b/attachment.html>

From matthieu.brucher at gmail.com  Sat Dec 29 03:36:20 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sat, 29 Dec 2007 09:36:20 +0100
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
In-Reply-To: <4776059b.03be100a.54cf.2160@mx.google.com>
References: <4776059b.03be100a.54cf.2160@mx.google.com>
Message-ID: <e76aa17f0712290036k7dc739c2r31be1a52ec647d8e@mail.gmail.com>

Hi,

The problem is the space in the name of the folder. It should not be there
(Python is usually not installed in Program Files). This problem arises
regularly with Linux too for inclusion paths with some tools.

Matthieu

2007/12/29, mani sabri <mani.sabri at gmail.com>:
>
>  Hi
>
> Happy holidays
>
> I'm using python 2.4.4 , scipy 0.6.0 , numpy 1.0.4  and mingw gcc 3.4.5 on
> win xp sp2
>
> When I try to run a weave example (e.x array3d.py) from "C:\PROGRAM
> FILES\Python24\Lib\site-packages\scipy\weave\examples" I have the following
> typical output:
>
>
>
> numpy:
>
> [[[ 0  1  2  3]
>
>   [ 4  5  6  7]
>
>   [ 8  9 10 11]]
>
>
>
>  [[12 13 14 15]
>
>   [16 17 18 19]
>
>   [20 21 22 23]]]
>
>
>
> Pure Inline:
>
> <weave: compiling>
>
> Found executable C:\MINGW\BIN\g++.exe
>
> g++.exe: C:\program: No such file or directory
>
> g++.exe: files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp:
> No such file or directory
>
> g++.exe: files\python24\lib\site-packages\scipy\weave\scxx\weave_imp.o: No
> such file or directory
>
> g++.exe: no input files
>
>
>
> And This trace back:
>
>
>
> CompileError: error: Command "g++ -mno-cygwin -O2 -Wall -I"C:\program
> files\Python24\lib\site-packages\scipy\weave" -I"C:\program
> files\Python24\lib\site-packages\scipy\weave\scxx" -I"C:\program
> files\Python24\lib\site-packages\numpy\core\include" -I"C:\program
> files\Python24\include" -I"C:\program files\Python24\PC" -c C:\program
> files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp -o
> c:\docume~1\man\locals~1\temp\man\python24_intermediate\compiler_12e837eb1ea3ab5199fbcc0e83015e3f\Release\program
> files\python24\lib\site-packages\scipy\weave\scxx\weave_imp.o" failed with
> exit status 1
>
>
>
> IMHO When I put quotes around ?c argument  in above command (-c
> "C:\program
> files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp") every
> thing is fine.
>
> Can anyone help me to hack this around?
>
>
>
> Cheers
>
> Mani.
>
>
>
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>


-- 
French PhD student
Website : http://matthieu-brucher.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/20071229/0c77e7ea/attachment.html>

From mani.sabri at gmail.com  Sat Dec 29 04:08:18 2007
From: mani.sabri at gmail.com (mani sabri)
Date: Sat, 29 Dec 2007 12:38:18 +0330
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
In-Reply-To: <e76aa17f0712290036k7dc739c2r31be1a52ec647d8e@mail.gmail.com>
Message-ID: <47760eda.1a7c100a.4a38.13e6@mx.google.com>

 

Hi

Yeah, exactly. do you know where is the code responsible for it? I wasn't
able to find the code for -c option because it does not exist in "-c " form
in any .py file in weave subdir. May be it is in makefile.in. imho this will
be solved by putting quotes around the file path. 

 

p.s: numpy.distutils.tests() runs without this problem from the same path.!
("C:\program files\Python24..")

 

  _____  

From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org] On
Behalf Of Matthieu Brucher
Sent: Saturday, December 29, 2007 12:06 PM
To: SciPy Users List
Subject: Re: [SciPy-user] weave and mingw on win xp sp2 bug?

 

Hi,

The problem is the space in the name of the folder. It should not be there
(Python is usually not installed in Program Files). This problem arises
regularly with Linux too for inclusion paths with some tools.

Matthieu

2007/12/29, mani sabri <mani.sabri at gmail.com>:

CompileError: error: Command "g++ -mno-cygwin -O2 -Wall -I"C:\program
files\Python24\lib\site-packages\scipy\weave" -I"C:\program
files\Python24\lib\site-packages\scipy\weave\scxx" -I"C:\program
files\Python24\lib\site-packages\numpy\core\include" -I"C:\program
files\Python24\include" -I"C:\program files\Python24\PC" -c C:\program
files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp -o
c:\docume~1\man\locals~1\temp\man\python24_intermediate\compiler_12e837eb1ea
3ab5199fbcc0e83015e3f\Release\program
files\python24\lib\site-packages\scipy\weave\scxx\weave_imp.o" failed with
exit status 1

IMHO When I put quotes around -c argument  in above command (-c "C:\program
files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp") every
thing is fine.

Can anyone help me to hack this around?

 

 


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




-- 
French PhD student
Website : http://matthieu-brucher.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/20071229/721f2a31/attachment.html>

From matthieu.brucher at gmail.com  Sat Dec 29 04:17:40 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sat, 29 Dec 2007 10:17:40 +0100
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
In-Reply-To: <47760eda.1a7c100a.4a38.13e6@mx.google.com>
References: <e76aa17f0712290036k7dc739c2r31be1a52ec647d8e@mail.gmail.com>
	<47760eda.1a7c100a.4a38.13e6@mx.google.com>
Message-ID: <e76aa17f0712290117p39674857l735a88438aa98d86@mail.gmail.com>

2007/12/29, mani sabri <mani.sabri at gmail.com>:
>
>
>
> Hi
>
> Yeah, exactly. do you know where is the code responsible for it? I wasn't
> able to find the code for ?c option because it does not exist in "-c " form
> in any .py file in weave subdir. May be it is in makefile.in. imho this
> will be solved by putting quotes around the file path.
>


It must be in Python distutils module, so your chances to fix it this way
are next to zero :(


p.s: numpy.distutils.tests() runs without this problem from the same path.!
> ("C:\program files\Python24?.")
>

That can happen.
Your best shot is using the default location for Python, that is
c:\Python24\ This way, it will work.

Matthieu
-- 
French PhD student
Website : http://matthieu-brucher.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/20071229/56e7ac25/attachment.html>

From mani.sabri at gmail.com  Sat Dec 29 04:52:32 2007
From: mani.sabri at gmail.com (mani sabri)
Date: Sat, 29 Dec 2007 13:22:32 +0330
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
In-Reply-To: <e76aa17f0712290117p39674857l735a88438aa98d86@mail.gmail.com>
Message-ID: <47761935.0ec5100a.5a49.15f2@mx.google.com>

>-----Original Message-----
>From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org] On
>Behalf Of Matthieu Brucher
>Sent: Saturday, December 29, 2007 12:48 PM
>
>2007/12/29, mani sabri <mani.sabri at gmail.com>:
>	Hi
>	Yeah, exactly. do you know where is the code responsible for it? I
>wasn't able to find the code for -c option because it does not exist in "-c
>" form in any .py file in weave subdir. May be it is in makefile.in. imho
>this will be solved by putting quotes around the file path.
>
>It must be in Python distutils module, so your chances to fix it this way
>are next to zero :(
>
>	p.s: numpy.distutils.tests() runs without this problem from the same
>path.! ("C:\program files\Python24..")
>
>That can happen.
>Your best shot is using the default location for Python, that is
>c:\Python24\ This way, it will work.

Then I should do that. Its time to attack the environment variables once
more! ;)

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



From grh at mur.at  Sat Dec 29 06:05:32 2007
From: grh at mur.at (Georg Holzmann)
Date: Sat, 29 Dec 2007 12:05:32 +0100
Subject: [SciPy-user] debugging scipy
Message-ID: <477629FC.5020306@mur.at>

Hallo!

I have a problem using scipy in the python debugger (I do that with the 
eric4 IDE).

When I step through the commands, I got the an error (only in debug 
mode!) at the statement:
"import scipy"
and the error is:
------8<--------
The file 
/usr/lib/python2.5/site-packages/setuptools-0.6c7-py2.5.egg/pkg_resources.py 
could not be opened.
------8<--------
and then:
------8<--------
The debugged program raised the exception OSError
"[Errno 2] No such file or directory: '/home/holzi/scipytest/test'"
File: 
/usr/lib/python2.5/site-packages/setuptools-0.6c7-py2.5.egg/pkg_resources.py, 
Line: 1659
Break here?
------8<--------

But my setuptools installation seems to be ok and the file 
/usr/lib/python2.5/site-packages/setuptools-0.6c7-py2.5.egg/ is present.
(my scipy version is 0.6.0, Python 2.5.1)

Has anyone a hint what I can try to resolve this problem ?

Many thanks,
LG
Georg


From mani.sabri at gmail.com  Sat Dec 29 07:04:56 2007
From: mani.sabri at gmail.com (mani sabri)
Date: Sat, 29 Dec 2007 15:34:56 +0330
Subject: [SciPy-user] weave.test() win mingw
Message-ID: <4776383d.0856100a.45fe.2e65@mx.google.com>

Hi 
Below is the result of weave.test() on win xp sp2. is it ok? Specially those
"error removing"s and "Permission denied". is it safe to use weave In this
condition?

Mani
---------------
>>>weave.test()
  Found 1/1 tests for scipy.weave.ast_tools
  Found 2/2 tests for scipy.weave.blitz_tools
  Found 9/9 tests for scipy.weave.build_tools
  Found 0/0 tests for scipy.weave.c_spec
  Found 26/26 tests for scipy.weave.catalog
building extensions here:
c:\docume~1\man\locals~1\temp\man\python24_compiled\m0
  Found 1/1 tests for scipy.weave.ext_tools
  Found 0/0 tests for scipy.weave.inline_tools
  Found 74/74 tests for scipy.weave.size_check
  Found 16/16 tests for scipy.weave.slice_handler
  Found 3/3 tests for scipy.weave.standard_array_spec
  Found 0/0 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
............................error removing
c:\docume~1\man\locals~1\temp\tmpcbvq4wcat_test:
c:\docume~1\man\locals~1\temp\tmpcbvq4wcat_test\win3224compiled_catalog:
Permission denied
error removing c:\docume~1\man\locals~1\temp\tmpcbvq4wcat_test:
c:\docume~1\man\locals~1\temp\tmpcbvq4wcat_test: Directory not empty
............................................................................
....................
----------------------------------------------------------------------
Ran 132 tests in 1.250s

OK
<unittest._TextTestResult run=132 errors=0 failures=0>
>>>



From odonnems at yahoo.com  Sat Dec 29 09:34:09 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Sat, 29 Dec 2007 06:34:09 -0800 (PST)
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
Message-ID: <32610.62959.qm@web58013.mail.re3.yahoo.com>

Hi sorry I did not reply about your inquiry of using weave. It took me quite a bit of time to get this to work and I never got any responses. Unfortunately, I am heading out the door so I will keep this extremely short until I can provide a more complete response tonight or tomorrow morning. I was able to get weave inline to work using python 2.5.1 and mingw 5.0.3. The path to this needs to be added to your pythonpath variable and you need to specify 'gcc' as your compiler. I also worked with VS2003 and VS2005 Express. And I worked with python 2.4.1 and 2.4.4. None of these combinations worked. There was also a problem in numpy with finding the python.exe file. I cannot remember if this error was given with VS compiler or both VS and MingW.

Anyhow, hopefully this will help you out and you can upgrade. I will provide all my notes on what I  tried and what I was able to get to work ASAP (tonight or tomorrow).


mike


----- Original Message ----
From: mani sabri <mani.sabri at gmail.com>
To: scipy-user at scipy.org
Sent: Saturday, December 29, 2007 1:28:51 AM
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?





 
 

<!--

 /* Style Definitions */
 p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;margin-bottom:.0001pt;font-size:12.0pt;font-family:"Times New Roman";}
a:link, span.MsoHyperlink
	{color:blue;text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{color:purple;text-decoration:underline;}
span.EmailStyle17
	{font-family:Arial;color:windowtext;}
 _filtered {margin:1.0in 1.25in 1.0in 1.25in;}
div.Section1
	{}
-->






Hi 
 

Happy holidays
 

I?m using python 2.4.4 , scipy 0.6.0 , numpy 1.0.4
 and mingw gcc 3.4.5 on win xp sp2 
 

When I try to run a weave example (e.x array3d.py) from
?C:\PROGRAM FILES\Python24\Lib\site-packages\scipy\weave\examples?
I have the following typical output:
 

  
 

numpy:
 

[[[ 0  1  2  3]
 

  [ 4  5  6  7]
 

  [ 8  9 10 11]]
 

  
 

 [[12 13 14 15]
 

  [16 17 18 19]
 

  [20 21 22 23]]]
 

  
 

Pure Inline:
 

<weave: compiling>
 

Found executable
C:\MINGW\BIN\g++.exe
 

g++.exe: C:\program: No such file or
directory
 

g++.exe: files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp:
No such file or directory
 

g++.exe:
files\python24\lib\site-packages\scipy\weave\scxx\weave_imp.o: No such file or
directory
 

g++.exe: no input files
 

  
 

And This trace back:
 

  
 

CompileError: error: Command "g++ -mno-cygwin -O2 -Wall
-I"C:\program files\Python24\lib\site-packages\scipy\weave"
-I"C:\program files\Python24\lib\site-packages\scipy\weave\scxx"
-I"C:\program files\Python24\lib\site-packages\numpy\core\include"
-I"C:\program files\Python24\include" -I"C:\program
files\Python24\PC" -c C:\program
files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp -o
c:\docume~1\man\locals~1\temp\man\python24_intermediate\compiler_12e837eb1ea3ab5199fbcc0e83015e3f\Release\program
files\python24\lib\site-packages\scipy\weave\scxx\weave_imp.o" failed with
exit status 1
 

  
 

IMHO When I put quotes around ?c argument  in
above command (-c ?C:\program
files\Python24\lib\site-packages\scipy\weave\scxx\weave_imp.cpp?) every
thing is fine.
 

Can anyone help me to hack this around?
 

  
 

Cheers
 

Mani.
 

  
 

  
 










      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071229/e19f175a/attachment.html>

From stefan at sun.ac.za  Sat Dec 29 11:43:25 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Sat, 29 Dec 2007 18:43:25 +0200
Subject: [SciPy-user] Scipy for Cluster Detection
In-Reply-To: <C8CF9489-1457-46AF-B453-62A338B422B5@cs.toronto.edu>
References: <4772F19C.6040308@gmail.com>
	<C8CF9489-1457-46AF-B453-62A338B422B5@cs.toronto.edu>
Message-ID: <20071229164325.GL9051@mentat.za.net>

On Thu, Dec 27, 2007 at 01:24:26AM -0500, David Warde-Farley wrote:
> Once you have this your problem is what graph theorists would call  
> identifying the "connected components", which is a well-studied  
> problem and can be done fairly easily, Google "finding connected  
> components" or pick up any algorithms textbook at your local library  
> such as the Cormen et al "Introduction to Algorithms".

I have implemented the O(N) algorithm described in

  Christophe Fiorio and Jens Gustedt,
  "Two linear time Union-Find strategies for image processing",
  Theoretical Computer Science 154 (1996), pp. 165-181.

and

  Kensheng Wu, Ekow Otoo and Arie Shoshani,
  "Optimizing connected component labeling algorithms",
  Paper LBNL-56864, 2005,
  Lawrence Berkeley National Laboratory
  (University of California),
  http://repositories.cdlib.org/lbnl/LBNL-56864.

The code can be found at

http://mentat.za.net/source/connected_components.tar.bz2

Run 'scons' to compile (I can send a setup.py if necessary).

Typical usage:

x = N.array([[0, 0, 3, 2, 1, 9],
             [0, 1, 1, 9, 2, 9],
	     [0, 0, 1, 9, 9, 9],
             [3, 1, 1, 5, 3, 0]])

labels = connected_components.linear(x)

Result:

[[0, 0, 1, 2, 3, 4],
 [0, 5, 5, 4, 6, 4],
 [0, 0, 5, 4, 4, 4],
 [7, 5, 5, 8, 9, 10]])

The given source searches for neighbours that have the same values,
but this can easily be modified to look for values within a certain
difference range.

Regards
St?fan


From corrada at cs.umass.edu  Sat Dec 29 12:37:52 2007
From: corrada at cs.umass.edu (Andres Corrada-Emmanuel)
Date: Sat, 29 Dec 2007 12:37:52 -0500
Subject: [SciPy-user] scipy.test segfaulting in Windows
Message-ID: <477685F0.1000506@cs.umass.edu>

Hello,

First off, let me thank all the developers of scipy for creating such an 
   excellent tool for scientific computation. After a week of struggling 
with some C++ packages and their python bindings, I found that the 
scipy.sandbox.delaunay package is just what I needed. It was a cinch to 
compile and install in Windows under Cygwin and natively.

I got greedy and wanted to make sure that the installation tests worked 
and got the segfault on both the native Python and the Cygwin Python 
sides that others have mentioned on the web.

I've tried building it with:

python setup.py build config_fc --noopt

but got the same segfault.

Any idea of how this can be fixed? What needs to be tweaked?

p.s.

Using UMFPACK in Cygwin requires that UMFPACK be linked to 
UFconfig/xerbla/libcerbla.a to avoid linker errors while building scipy 
with umfpack support

-- 
Andres Corrada-Emmanuel
Research Fellow
Aerial Imaging and Remote Sensing Lab
Computer Science Department
University of Massachusetts at Amherst
Blog: www.corrada.com/blog



From humufr at yahoo.fr  Sat Dec 29 12:47:02 2007
From: humufr at yahoo.fr (humufr at yahoo.fr)
Date: Sat, 29 Dec 2007 12:47:02 -0500
Subject: [SciPy-user] maskedarray moved?
In-Reply-To: <c7009a550712281303k5a134712kc4580e46bd224650@mail.gmail.com>
References: <200712281546.33014.humufr@yahoo.fr>
	<c7009a550712281303k5a134712kc4580e46bd224650@mail.gmail.com>
Message-ID: <200712291247.03322.humufr@yahoo.fr>

Cool great news, I was using only maskedarray for sometime now so if it'll be 
include in numpy that will be far better.

Thanks,

N.

Le Friday 28 December 2007 16:03:37 Jarrod Millman, vous avez ?crit?:
> It is now in a numpy branch:
> http://projects.scipy.org/scipy/numpy/browser/branches/maskedarray
>
> and will be merged with the numpy trunk soon.  You can view the API
> changes with the existing numpy.ma here:
> http://svn.scipy.org/svn/numpy/branches/maskedarray/numpy/ma/API_CHANGES.tx
>t
>
> Thanks,
> Jarrod
>
> On Dec 28, 2007 12:46 PM,  <humufr at yahoo.fr> wrote:
> >                 Hi,
> >
> > I would like to know where is maskedarray. It's not anymore inside the
> > sandbox? Where can I find it now?
> >
> > Thanks,
> >
> > N.
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user


From lists.steve at arachnedesign.net  Sat Dec 29 16:16:43 2007
From: lists.steve at arachnedesign.net (Steve Lianoglou)
Date: Sat, 29 Dec 2007 16:16:43 -0500
Subject: [SciPy-user] slicot and numpy/scipy
Message-ID: <205B1B5C-DFE5-4676-B3FC-343529818F9A@arachnedesign.net>

Hi all,

I saw some talk in the mail archives (between Travis and Nils) about  
interfacing scipy w/ slicot, but never saw anything come out of it.

I was just curious if this was ever seen through, or if there were  
problems that couldn't be worked around (aside from the inability to  
distribute slicot).

Thanks,
-steve


From robert.kern at gmail.com  Sat Dec 29 16:29:11 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Sat, 29 Dec 2007 16:29:11 -0500
Subject: [SciPy-user] debugging scipy
In-Reply-To: <477629FC.5020306@mur.at>
References: <477629FC.5020306@mur.at>
Message-ID: <4776BC27.3050304@gmail.com>

Georg Holzmann wrote:
> Hallo!
> 
> I have a problem using scipy in the python debugger (I do that with the 
> eric4 IDE).
> 
> When I step through the commands, I got the an error (only in debug 
> mode!) at the statement:
> "import scipy"
> and the error is:
> ------8<--------
> The file 
> /usr/lib/python2.5/site-packages/setuptools-0.6c7-py2.5.egg/pkg_resources.py 
> could not be opened.
> ------8<--------
> and then:
> ------8<--------
> The debugged program raised the exception OSError
> "[Errno 2] No such file or directory: '/home/holzi/scipytest/test'"
> File: 
> /usr/lib/python2.5/site-packages/setuptools-0.6c7-py2.5.egg/pkg_resources.py, 
> Line: 1659
> Break here?
> ------8<--------
> 
> But my setuptools installation seems to be ok and the file 
> /usr/lib/python2.5/site-packages/setuptools-0.6c7-py2.5.egg/ is present.
> (my scipy version is 0.6.0, Python 2.5.1)
> 
> Has anyone a hint what I can try to resolve this problem ?

Eric apparently does not know anything about zipped eggs. The setuptools egg is
zipped and pkg_resource.py is inside of it. In debug mode, I imagine it tries to
load the source of every module that gets imported.

Remove the lines in scipy/__init__.py:

"""
try:
    import pkg_resources as _pr # activate namespace packages (manipulates __path__)
    del _pr
except ImportError:
    pass
"""

I'm not really sure why we put them there, though.

-- 
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 corrada at cs.umass.edu  Sat Dec 29 20:26:14 2007
From: corrada at cs.umass.edu (Andres Corrada-Emmanuel)
Date: Sat, 29 Dec 2007 20:26:14 -0500
Subject: [SciPy-user] scipy.test segfaulting in Windows
In-Reply-To: <477685F0.1000506@cs.umass.edu>
References: <477685F0.1000506@cs.umass.edu>
Message-ID: <4776F3B6.4050008@cs.umass.edu>

The problem is related to the test_histogram tests in test_ndimage.py. 
Commenting out the tests results in the test suite completing but two 
failures occur later on (with no segfault) also related to nd_image:

generic filter 1
generic 1d filter 1

Andres Corrada-Emmanuel wrote:
> Hello,
> 
> First off, let me thank all the developers of scipy for creating such an 
>    excellent tool for scientific computation. After a week of struggling 
> with some C++ packages and their python bindings, I found that the 
> scipy.sandbox.delaunay package is just what I needed. It was a cinch to 
> compile and install in Windows under Cygwin and natively.
> 
> I got greedy and wanted to make sure that the installation tests worked 
> and got the segfault on both the native Python and the Cygwin Python 
> sides that others have mentioned on the web.
> 
> I've tried building it with:
> 
> python setup.py build config_fc --noopt
> 
> but got the same segfault.
> 
> Any idea of how this can be fixed? What needs to be tweaked?
> 
> p.s.
> 
> Using UMFPACK in Cygwin requires that UMFPACK be linked to 
> UFconfig/xerbla/libcerbla.a to avoid linker errors while building scipy 
> with umfpack support
> 

-- 
Andres Corrada-Emmanuel
Research Fellow
Aerial Imaging and Remote Sensing Lab
Computer Science Department
University of Massachusetts at Amherst
Blog: www.corrada.com/blog



From stefan at sun.ac.za  Sat Dec 29 21:06:38 2007
From: stefan at sun.ac.za (Stefan van der Walt)
Date: Sun, 30 Dec 2007 04:06:38 +0200
Subject: [SciPy-user] scipy.test segfaulting in Windows
In-Reply-To: <4776F3B6.4050008@cs.umass.edu>
References: <477685F0.1000506@cs.umass.edu> <4776F3B6.4050008@cs.umass.edu>
Message-ID: <20071230020638.GA11137@mentat.za.net>

Hi Andres

On Sat, Dec 29, 2007 at 08:26:14PM -0500, Andres Corrada-Emmanuel wrote:
> The problem is related to the test_histogram tests in test_ndimage.py. 
> Commenting out the tests results in the test suite completing but two 
> failures occur later on (with no segfault) also related to nd_image:
> 
> generic filter 1
> generic 1d filter 1

This has been fixed in SVN for some time, so try that version instead.

Regards
St?fan


From fperez.net at gmail.com  Sat Dec 29 22:08:55 2007
From: fperez.net at gmail.com (Fernando Perez)
Date: Sat, 29 Dec 2007 20:08:55 -0700
Subject: [SciPy-user] Bundling numpy/scipy with applications
In-Reply-To: <473953CC.7010402@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>
	<47394E44.6010608@slac.stanford.edu>
	<473953CC.7010402@ar.media.kyoto-u.ac.jp>
Message-ID: <db6b5ecc0712291908v4a9335a3s12708a33c5434b7a@mail.gmail.com>

On Nov 13, 2007 12:35 AM, David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
> 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).

We finally tracked this down to a system-wide numpy confusing the
brittle hack that __init__ was using to detect being run from the
source directory.

This commit:

http://projects.scipy.org/scipy/numpy/changeset/4663

Fixes it and allows you to do an in-place build of numpy via

python setup.py build_src --inplace
python setup.py build_ext --inplace

for development.

It's a bit of an ugly hack, but at least it works robustly and avoids
the problem above (which bit me today).

Please let me know if the approach in the patch causes problems for
any other types of usage.

Cheers,

f


From corrada at cs.umass.edu  Sat Dec 29 23:03:58 2007
From: corrada at cs.umass.edu (Andres Corrada-Emmanuel)
Date: Sat, 29 Dec 2007 23:03:58 -0500
Subject: [SciPy-user] scipy.test segfaulting in Windows
In-Reply-To: <20071230020638.GA11137@mentat.za.net>
References: <477685F0.1000506@cs.umass.edu> <4776F3B6.4050008@cs.umass.edu>
	<20071230020638.GA11137@mentat.za.net>
Message-ID: <477718AE.8050408@cs.umass.edu>

Thanks Stefan, using the svn versions of nd_image.[ch] fixed the problem.

Stefan van der Walt wrote:
> Hi Andres
> 
> On Sat, Dec 29, 2007 at 08:26:14PM -0500, Andres Corrada-Emmanuel wrote:
>> The problem is related to the test_histogram tests in test_ndimage.py. 
>> Commenting out the tests results in the test suite completing but two 
>> failures occur later on (with no segfault) also related to nd_image:
>>
>> generic filter 1
>> generic 1d filter 1
> 
> This has been fixed in SVN for some time, so try that version instead.
> 
> Regards
> St?fan
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

-- 
Andres Corrada-Emmanuel
Research Fellow
Aerial Imaging and Remote Sensing Lab
Computer Science Department
University of Massachusetts at Amherst
Blog: www.corrada.com/blog



From nwagner at iam.uni-stuttgart.de  Sun Dec 30 03:26:05 2007
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sun, 30 Dec 2007 09:26:05 +0100
Subject: [SciPy-user] slicot and numpy/scipy
In-Reply-To: <205B1B5C-DFE5-4676-B3FC-343529818F9A@arachnedesign.net>
References: <205B1B5C-DFE5-4676-B3FC-343529818F9A@arachnedesign.net>
Message-ID: <web-94042815@uni-stuttgart.de>

On Sat, 29 Dec 2007 16:16:43 -0500
  Steve Lianoglou <lists.steve at arachnedesign.net> wrote:
> Hi all,
> 
> I saw some talk in the mail archives (between Travis and 
>Nils) about  
> interfacing scipy w/ slicot, but never saw anything come 
>out of it.
> 
> I was just curious if this was ever seen through, or if 
>there were  
> problems that couldn't be worked around (aside from the 
>inability to  
> distribute slicot).
> 
> Thanks,
> -steve
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

  
Hi Steve,

IIRC it was a license problem but now we have scikits :-).

http://www.slicot.org/

Commercial users need to reach a license agreement before 
using SLICOT software

http://www.slicot.org/index.php?site=access

Cheers,
           Nils


From mani.sabri at gmail.com  Sun Dec 30 04:34:18 2007
From: mani.sabri at gmail.com (mani sabri)
Date: Sun, 30 Dec 2007 13:04:18 +0330
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
In-Reply-To: <32610.62959.qm@web58013.mail.re3.yahoo.com>
Message-ID: <4777666e.0ac0100a.3bee.7fff@mx.google.com>

>-----Original Message-----
>From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org] On
>Behalf Of Michael ODonnell
>Sent: Saturday, December 29, 2007 6:04 PM
>To: SciPy Users List
>
>Hi sorry I did not reply about your inquiry of using weave. It took me
>quite a bit of time to get this to work and I never got any responses.
>Unfortunately, I am heading out the door so I will keep this extremely
>short until I can provide a more complete response tonight or tomorrow
>morning. I was able to get weave inline to work using python 2.5.1 and
>mingw 5.0.3. The path to this needs to be added to your pythonpath variable
>and you need to specify 'gcc' as your compiler. I also worked with VS2003
>and VS2005 Express. And I worked with python 2.4.1 and 2.4.4. None of these
>combinations worked. There was also a problem in numpy with finding the
>python.exe file. I cannot remember if this error was given with VS compiler
>or both VS and MingW.
>

I managed to run/compile my Time Wrapping function with weave and python
2.4.4 an hour ago. Although the weave.tests continues to show ugly
"Permission Denied" message I had no problem with my func. (and like most
windows users I don't care about that message! Its windows! ;) )

As I remember it is the python24.a file in 2.4.4 that makes the whole
difference because gcc don't know anything about Microsoft lib files.
Anyhow here is a brief state of my system:

- scipy 0.6.0
- numpy 1.0.4
- Python in c:\python24 (no spaces in path! Its important)
- environment variables:
	python_lib = c:\Python24\libs\python24.a
	pythonpath = c:\Python24;C:\PROGRAM FILES\Pyrex-0.9.6.4;
	swig_lib =	C:\PROGRAM FILES\swingwin-1.3.27\python
	python_include = C:\Python24\include;
	path = C:\MINGW\BIN;C:\CYGWIN\BIN; C:\PROGRAM FILES\swingwin-
1.3.27\python;c:\python24

p.s: I saw some posts about cleaning up weave.tests in scipy-dev these days.
I hope they take a look at weave.example directory. 



From grh at mur.at  Sun Dec 30 08:37:38 2007
From: grh at mur.at (Georg Holzmann)
Date: Sun, 30 Dec 2007 14:37:38 +0100
Subject: [SciPy-user] debugging scipy
In-Reply-To: <4776BC27.3050304@gmail.com>
References: <477629FC.5020306@mur.at> <4776BC27.3050304@gmail.com>
Message-ID: <47779F22.7020105@mur.at>

Hallo!

> Eric apparently does not know anything about zipped eggs. The setuptools egg is
> zipped and pkg_resource.py is inside of it. In debug mode, I imagine it tries to
> load the source of every module that gets imported.

Thanks - yes sorry, this problem is only with eric. I just tried 
pydev/eclipse and there is no such problem.
I wanted to find a good IDE for numpy/scipy scripts - pydev/eclipse 
works quite good, but there is no interactive console during debugging 
where you could plot some numarrays ...

Thanks,
LG
Georg


From corrada at cs.umass.edu  Sun Dec 30 08:46:46 2007
From: corrada at cs.umass.edu (Andres Corrada-Emmanuel)
Date: Sun, 30 Dec 2007 08:46:46 -0500
Subject: [SciPy-user] debugging scipy
In-Reply-To: <47779F22.7020105@mur.at>
References: <477629FC.5020306@mur.at> <4776BC27.3050304@gmail.com>
	<47779F22.7020105@mur.at>
Message-ID: <4777A146.1040309@cs.umass.edu>

Hello,

On a different slant on this topic: Is there documentation on how to 
setup a debugging environment for scipy? I mean by that some sort of 
step-by-step instructions on how to build a debugging version, what 
tools are useful (nm, objdump, etc.) and how to step thru the pdb and/or 
gdb at the same time. I once knew how to do this with extension modules 
with gdb but have lost the knowledge after years of disuse. This sort of 
documentation could help people like me who are users of scipy to 
acquire skills that can help the scipy developers when debugging problems.

Georg Holzmann wrote:
> Hallo!
> 
>> Eric apparently does not know anything about zipped eggs. The setuptools egg is
>> zipped and pkg_resource.py is inside of it. In debug mode, I imagine it tries to
>> load the source of every module that gets imported.
> 
> Thanks - yes sorry, this problem is only with eric. I just tried 
> pydev/eclipse and there is no such problem.
> I wanted to find a good IDE for numpy/scipy scripts - pydev/eclipse 
> works quite good, but there is no interactive console during debugging 
> where you could plot some numarrays ...
> 
> Thanks,
> LG
> Georg
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

-- 
Andres Corrada-Emmanuel
Research Fellow
Aerial Imaging and Remote Sensing Lab
Computer Science Department
University of Massachusetts at Amherst
Blog: www.corrada.com/blog



From grh at mur.at  Sun Dec 30 10:25:42 2007
From: grh at mur.at (Georg Holzmann)
Date: Sun, 30 Dec 2007 16:25:42 +0100
Subject: [SciPy-user] debugging scipy
In-Reply-To: <4777A146.1040309@cs.umass.edu>
References: <477629FC.5020306@mur.at>
	<4776BC27.3050304@gmail.com>	<47779F22.7020105@mur.at>
	<4777A146.1040309@cs.umass.edu>
Message-ID: <4777B876.9060102@mur.at>

Hallo!

> On a different slant on this topic: Is there documentation on how to 
> setup a debugging environment for scipy? I mean by that some sort of 
> step-by-step instructions on how to build a debugging version, what 
> tools are useful (nm, objdump, etc.) and how to step thru the pdb and/or 
> gdb at the same time. I once knew how to do this with extension modules 

Yes this would be nice - I am currently also trying to debug python and 
c++ modules for python - so far I tried this now with eclipse (with CDT 
and PYDEV plugin) but I am not yet there where I want to be ... ;) ... 
however, I will try again next year ...

LG
Georg


From robert.kern at gmail.com  Sun Dec 30 16:39:50 2007
From: robert.kern at gmail.com (Robert Kern)
Date: Sun, 30 Dec 2007 16:39:50 -0500
Subject: [SciPy-user] slicot and numpy/scipy
In-Reply-To: <web-94042815@uni-stuttgart.de>
References: <205B1B5C-DFE5-4676-B3FC-343529818F9A@arachnedesign.net>
	<web-94042815@uni-stuttgart.de>
Message-ID: <47781026.2060901@gmail.com>

Nils Wagner wrote:
> IIRC it was a license problem but now we have scikits :-).

We also have a policy that scikits must have an open source license. SLICOT is
distinctly not open source. So no, SLICOT wrappers will not be a part of scipy
or scikits. You're welcome to write wrappers hosted elsewhere and discuss them
here, though.

-- 
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 odonnems at yahoo.com  Sun Dec 30 18:01:22 2007
From: odonnems at yahoo.com (Michael ODonnell)
Date: Sun, 30 Dec 2007 15:01:22 -0800 (PST)
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
Message-ID: <359774.29886.qm@web58014.mail.re3.yahoo.com>

Thanks for the info on this. Sorry I was slow at responding to your email. I did compile my notes as to how I got weave.inline to work. I know that when I was working on this before and trying a million different combinations, I also changed the location of my temp directory in the environment variables to C:/Temp. I have learned that everything tends to work best when there are no spaces in directories and ISO standards are followed. I will go ahead and post my notes on what I did but it sounds like you have everything working so feel free to ignore.



Like I mentioned I was only
able to get weave.inline to work with python 2.5.1 and mingw 5.0.3. I tried
python 2.4.1 and 2.4.4 with visual studio compilers and mingw. I hope this
information is of some help. This process was new for me so it may not be the
best solution or accurate but it does work for me.


 


OS: Windows XP SP2
professional


Processor: x86


 


I suggest you check that MinGW
is installed correctly by following the directions on their website. A summary
of these directions include: First install MSYS, say, to directory C:\msys\1.0.
Then install MinGW to C:\msys\1.0\mingw directory. If you choose to install
MinGW to C:\mingw as the installer suggests by default then you have to create
a file C:\msys\1.0\etc\fstab containing c:/mingw /mingw and then restart your
computer so that /mingw will be mounted. Once this is done you might test the
compiler on something to eliminate this as a source of error. Then make sure
you add the MinGW path to PYTHONPATH environment variable (e.g., % PYTHONPATH%;
C:\msys\1.0\bin;C:\msys\1.0\mingw\bin;).


 


I then had to modify the numpy
distutils exec_command.py. This is likely found in
C:\Python25\Lib\site-packages\numpy\ distutils:


 


                    I commented out this code starting on line 67 or
so:


    '''


    if os.name
in ['nt','dos']:


       
fdir,fn = os.path.split(pythonexe)


        fn =
fn.upper().replace('PYTHONW','PYTHON')


       
pythonexe = os.path.join(fdir,fn)


        assert
os.path.isfile(pythonexe), '%r is not a file' % (pythonexe,)


    '''


        And then I replaced with these lines:


        ###MOD edits


        #C:\Python24\python.exe


        import string


        exe = 'python.exe'


        for p in
string.split(os.environ['Path'],';'):


            fn =
os.path.join(os.path.abspath(p),exe)


            if os.path.isfile(fn):


                pythonexe = fn


        ###End of MOD edits


 


 


These
edits worked for me.


 


I
was never successful in using VS2003 toolkit or VS2005 Express. I was able to
install these and compile some python modules successfully so they did work as
compilers. However, changes to the SDK had to be implemented in order for these
to work as well. I am not sure why I could not get weave to work with these
compilers but once I got something to work I gave up trying to understand due
to my time constraints. The alterations I made to the compilers are included
below. Apparently the problem has to do with Microsoft and Express. The full software
package of visual studio apparently does not have these problems. Here are the
edits I made to get these to compile python modules (weave still did not work
however).


 


Source for this info: http://www.codeproject.com/KB/wtl/WTLExpress.aspx


 


 


1. cd C:\Program
Files\Microsoft Platform SDK\include\atl


 


Change SetChainEntry function
at line 1725 of atlwin.h - define "int i" at the first line of the
function body.


BOOL SetChainEntry(DWORD
dwChainID, CMessageMap* pObject, DWORD


dwMsgMapID = 0)


{


    int i;


    // first search for an existing entry


 


    


    for(i = 0; i < m_aChainEntry.GetSize();
i++)


 


 


2.Change AllocStdCallThunk
and FreeStdCallThunk at line 287 of atlbase.h to the new macros


/* Comment it


PVOID __stdcall __AllocStdCallThunk(VOID);


VOID __stdcall
__FreeStdCallThunk(PVOID);


 


#define AllocStdCallThunk()
__AllocStdCallThunk()


#define FreeStdCallThunk(p)
__FreeStdCallThunk(p)


 


#pragma comment(lib,
"atlthunk.lib")


*/


#define AllocStdCallThunk()
HeapAlloc(GetProcessHeap(), 


                              0,
sizeof(_stdcallthunk))


#define FreeStdCallThunk(p)
HeapFree(GetProcessHeap(), 0, p)


 


 


3. Download and install WTL
from SourceForge


cd  to WTL\AppWiz folder, double click
setup80x.js (or appropriate version to install the WTL Wizard into VC Express








----- Original Message ----
From: mani sabri <mani.sabri at gmail.com>
To: SciPy Users List <scipy-user at scipy.org>
Sent: Sunday, December 30, 2007 2:34:18 AM
Subject: Re: [SciPy-user] weave and mingw on win xp sp2 bug?


>-----Original Message-----
>From: scipy-user-bounces at scipy.org
 [mailto:scipy-user-bounces at scipy.org] On
>Behalf Of Michael ODonnell
>Sent: Saturday, December 29, 2007 6:04 PM
>To: SciPy Users List
>
>Hi sorry I did not reply about your inquiry of using weave. It took me
>quite a bit of time to get this to work and I never got any responses.
>Unfortunately, I am heading out the door so I will keep this extremely
>short until I can provide a more complete response tonight or tomorrow
>morning. I was able to get weave inline to work using python 2.5.1 and
>mingw 5.0.3. The path to this needs to be added to your pythonpath
 variable
>and you need to specify 'gcc' as your compiler. I also worked with
 VS2003
>and VS2005 Express. And I worked with python 2.4.1 and 2.4.4. None of
 these
>combinations worked. There was also a problem in numpy with finding
 the
>python.exe file. I cannot remember if this error was given with VS
 compiler
>or both VS and MingW.
>

I managed to run/compile my Time Wrapping function with weave and
 python
2.4.4 an hour ago. Although the weave.tests continues to show ugly
"Permission Denied" message I had no problem with my func. (and like
 most
windows users I don't care about that message! Its windows! ;) )

As I remember it is the python24.a file in 2.4.4 that makes the whole
difference because gcc don't know anything about Microsoft lib files.
Anyhow here is a brief state of my system:

- scipy 0.6.0
- numpy 1.0.4
- Python in c:\python24 (no spaces in path! Its important)
- environment variables:
    python_lib = c:\Python24\libs\python24.a
    pythonpath = c:\Python24;C:\PROGRAM FILES\Pyrex-0.9.6.4;
    swig_lib =    C:\PROGRAM FILES\swingwin-1.3.27\python
    python_include = C:\Python24\include;
    path = C:\MINGW\BIN;C:\CYGWIN\BIN; C:\PROGRAM FILES\swingwin-
1.3.27\python;c:\python24

p.s: I saw some posts about cleaning up weave.tests in scipy-dev these
 days.
I hope they take a look at weave.example directory. 

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






      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20071230/482523e1/attachment.html>

From mani.sabri at gmail.com  Mon Dec 31 04:08:47 2007
From: mani.sabri at gmail.com (mani sabri)
Date: Mon, 31 Dec 2007 12:38:47 +0330
Subject: [SciPy-user] weave and mingw on win xp sp2 bug?
In-Reply-To: <359774.29886.qm@web58014.mail.re3.yahoo.com>
Message-ID: <4778b1f7.05a4100a.100a.ffffd2d0@mx.google.com>

I just installed mingw in c:\mingw and moved my python24 from "c:\program
files\python24" to "c:\python24" and added those environment variables that
I mentioned before and every thing did not worked!(such as weave.test() and
some of most of examples but its not important because imho they are
outdated) but my code compiled successfully. 

>-----Original Message-----
>From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org] On
>Behalf Of Michael ODonnell
>Sent: Monday, December 31, 2007 2:31 AM
>To: SciPy Users List
>Subject: Re: [SciPy-user] weave and mingw on win xp sp2 bug?
>
>I also changed the location of my temp directory in the
>environment variables to C:/Temp. I have learned that everything tends to
>work best when there are no spaces in directories and ISO standards are
>followed. 
>
>Like I mentioned I was only able to get weave.inline to work with python
>2.5.1 and mingw 5.0.3. I tried python 2.4.1 and 2.4.4 with visual studio
>compilers and mingw. 
>
>OS: Windows XP SP2 professional
>Processor: x86
>
>I suggest you check that MinGW is installed correctly by following the
>directions on their website. A summary of these directions include: First
>install MSYS, say, to directory C:\msys\1.0. Then install MinGW to
>C:\msys\1.0\mingw directory. If you choose to install MinGW to C:\mingw as
>the installer suggests by default then you have to create a file
>C:\msys\1.0\etc\fstab containing c:/mingw /mingw and then restart your
>computer so that /mingw will be mounted. Once this is done you might test
>the compiler on something to eliminate this as a source of error. Then make
>sure you add the MinGW path to PYTHONPATH environment variable (e.g., %
>PYTHONPATH%; C:\msys\1.0\bin;C:\msys\1.0\mingw\bin;).
>
>
>I then had to modify the numpy distutils exec_command.py. This is likely
>found in C:\Python25\Lib\site-packages\numpy\ distutils:
>
>                    I commented out this code starting on line 67 or so:
>    '''
>    if os.name in ['nt','dos']:
>        fdir,fn = os.path.split(pythonexe)
>        fn = fn.upper().replace('PYTHONW','PYTHON')
>        pythonexe = os.path.join(fdir,fn)
>        assert os.path.isfile(pythonexe), '%r is not a file' % (pythonexe,)
>    '''
>
>        And then I replaced with these lines:
>        ###MOD edits
>        #C:\Python24\python.exe
>        import string
>        exe = 'python.exe'
>        for p in string.split(os.environ['Path'],';'):
>            fn = os.path.join(os.path.abspath(p),exe)
>            if os.path.isfile(fn):
>                pythonexe = fn
>        ###End of MOD edits
>
>These edits worked for me.
>
	It's strange. 

>
>I was never successful in using VS2003 toolkit or VS2005 Express. I was
>able to install these and compile some python modules successfully so they
>did work as compilers. However, changes to the SDK had to be implemented in
>order for these to work as well. I am not sure why I could not get weave to
>work with these compilers but once I got something to work I gave up trying
>to understand due to my time constraints. The alterations I made to the
>compilers are included below. Apparently the problem has to do with
>Microsoft and Express. The full software package of visual studio
>apparently does not have these problems. Here are the edits I made to get
>these to compile python modules (weave still did not work however).
>
>Source for this info: http://www.codeproject.com/KB/wtl/WTLExpress.aspx
>
>1. cd C:\Program Files\Microsoft Platform SDK\include\atl
>
>Change SetChainEntry function at line 1725 of atlwin.h - define "int i" at
>the first line of the function body.
>
>BOOL SetChainEntry(DWORD dwChainID, CMessageMap* pObject, DWORD
>
>dwMsgMapID = 0)
>{
>    int i;
>    // first search for an existing entry
>    for(i = 0; i < m_aChainEntry.GetSize(); i++)
>
>2.Change AllocStdCallThunk and FreeStdCallThunk at line 287 of atlbase.h to
>the new macros
>
>/* Comment it
>PVOID __stdcall __AllocStdCallThunk(VOID);
>VOID __stdcall __FreeStdCallThunk(PVOID);
>
>#define AllocStdCallThunk() __AllocStdCallThunk()
>
>#define FreeStdCallThunk(p) __FreeStdCallThunk(p)
>
>#pragma comment(lib, "atlthunk.lib")
>*/
>#define AllocStdCallThunk() HeapAlloc(GetProcessHeap(),
>                              0, sizeof(_stdcallthunk))
>#define FreeStdCallThunk(p) HeapFree(GetProcessHeap(), 0, p)
>
>3. Download and install WTL from SourceForge
>
>cd  to WTL\AppWiz folder, double click setup80x.js (or appropriate version
>to install the WTL Wizard into VC Express



From lev at columbia.edu  Mon Dec 31 13:07:51 2007
From: lev at columbia.edu (Lev Givon)
Date: Mon, 31 Dec 2007 13:07:51 -0500
Subject: [SciPy-user] Ann: OpenOpt v 0.15 (free optimization framework)
In-Reply-To: <47643642.80808@scipy.org>
References: <47643642.80808@scipy.org>
Message-ID: <20071231180751.GB26755@localhost.cc.columbia.edu>

Received from dmitrey on Sat, Dec 15, 2007 at 03:17:06PM EST:
> Hi all,
> we are glad to inform you about OpenOpt v 0.15 (release), free (license: 
> BSD) optimization framework for Python language programmers
> 
> Changes since previous release (September 4):
> 
>     * some new classes
>     * several new solvers written
>     * some more solvers connected
>     * NLP/NSP solver ralg can handle constrained problems
>     * some bugfixes
>     * some enhancements in graphical output (especially for constrained 
> problems)
> 
> Regards,
> OpenOpt developers
> http://scipy.org/scipy/scikits/wiki/OpenOpt
> http://openopt.blogspot.com/

Why are there two tarballs available for download on 
http://scipy.org/scipy/scikits/wiki/OpenOptInstall? They both seem to
contain the same version of the software.

							L.G.


From millman at berkeley.edu  Mon Dec 31 17:43:37 2007
From: millman at berkeley.edu (Jarrod Millman)
Date: Mon, 31 Dec 2007 14:43:37 -0800
Subject: [SciPy-user] planet.scipy.org
Message-ID: <c7009a550712311443r6c1e70a6q5afaaba2a08bcf95@mail.gmail.com>

Hey,

I just wanted to announce that we now have a NumPy/SciPy blog
aggregator thanks to Ga?l Varoquaux:  http://planet.scipy.org/

Feel free to contact me if you have a blog that you would like included.

Thanks,

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