From cournape at gmail.com  Mon Sep  1 03:13:51 2008
From: cournape at gmail.com (David Cournapeau)
Date: Mon, 1 Sep 2008 16:13:51 +0900
Subject: [SciPy-user] Create a spectrogram from a waveform
In-Reply-To: <86f16dc10808311352k32bebbf7qd67b77b936b6973b@mail.gmail.com>
References: <86f16dc10808301523x37dfe309y29d496564c6cb305@mail.gmail.com>
	<20080830202812.bpnp4vvz0g4g4o80@mail.enthought.com>
	<86f16dc10808310649v127c77acha3d13c5d03773a44@mail.gmail.com>
	<86f16dc10808311158h68c550edh1c11a55d43641378@mail.gmail.com>
	<e76aa17f0808311219x5dae498co8370a2d2a0680069@mail.gmail.com>
	<86f16dc10808311352k32bebbf7qd67b77b936b6973b@mail.gmail.com>
Message-ID: <5b8d13220809010013k5e54ceb0sfcd5460b059ecb0c@mail.gmail.com>

On Mon, Sep 1, 2008 at 5:52 AM, Ed McCaffrey <ed at edmccaffrey.net> wrote:
> Thanks for the reply.  I had not heard of audiolab before, but I just tried
> using it.
>
> Looking at audiolab made me realize that I had forgotten how a .wav stores
> the data for multiple channels, so that was why the spectrogram I generated
> before looked so odd.

You should not have to care how it is stored, normally. audiolab gives
you one column per channel; audiolab is just a wrapper around sndfile,
which handles interleaving/deinterleaving internally if necessary.

Also, this is not well documented, unfortunately, but if you don:t
have advanced needs, you can use the high level API ala matlab:

from scikits.audiolab import wavread

If you want to compute the spectrogram without matplotlib, this is not
too difficult: a spectrogram is a short time fourier, that is fourier
transform computed on windowed parts of your signal

http://en.wikipedia.org/wiki/Short-time_Fourier_transform

I hope to include a tool for automatically segmenting a signal into a
matrix of overlapping windows (implemented by A. Archibald) into numpy
after 1.2 release. With this, a spectrogram is 2-3 lines away in pure
numpy

cheers,

David


From luca.ciciriello at email.it  Mon Sep  1 05:56:28 2008
From: luca.ciciriello at email.it (luca.ciciriello at email.it)
Date: Mon, 1 Sep 2008 11:56:28 +0200
Subject: [SciPy-user] install SciPy on Mac OS X 10.4
Message-ID: <8687ec0214b346df50502c9af4a8f9b4@85.18.140.153>


Hi All.
I've installed NumPy 1.1.1 and now I want to install SciPy 0.6.0 on My Mac
OS X 
(10.4 PPC). I've got scipy-0.6.0.tar.gz.tar from SciPy site. On My system
is installed Xcode 2.5, so my default compiler is gcc 4.0.1. I've read in
INSTALL.TXT in SciPy distribution that SciPy is not so compatible with this
compiler, then to try to digit gcc_select 3.3. 
My question is: where have I to digit this gcc_select 3.3? 

Thanks in advance for any answer.

Luca 
 --
 Email.it, the professional e-mail, gratis per te: http://www.email.it/f
 
 Sponsor:
 Realizza i tuoi sogni con Carta Eureka. Fido fino a 3.000 euro, rate a
partire da 20 euro e canone gratis il 1? anno. Scoprila!
 Clicca qui: http://adv.email.it/cgi-bin/foclick.cgi?mid=7877&d=20080901




From ed at edmccaffrey.net  Mon Sep  1 07:20:12 2008
From: ed at edmccaffrey.net (Ed McCaffrey)
Date: Mon, 1 Sep 2008 07:20:12 -0400
Subject: [SciPy-user] Create a spectrogram from a waveform
In-Reply-To: <5b8d13220809010013k5e54ceb0sfcd5460b059ecb0c@mail.gmail.com>
References: <86f16dc10808301523x37dfe309y29d496564c6cb305@mail.gmail.com>
	<20080830202812.bpnp4vvz0g4g4o80@mail.enthought.com>
	<86f16dc10808310649v127c77acha3d13c5d03773a44@mail.gmail.com>
	<86f16dc10808311158h68c550edh1c11a55d43641378@mail.gmail.com>
	<e76aa17f0808311219x5dae498co8370a2d2a0680069@mail.gmail.com>
	<86f16dc10808311352k32bebbf7qd67b77b936b6973b@mail.gmail.com>
	<5b8d13220809010013k5e54ceb0sfcd5460b059ecb0c@mail.gmail.com>
Message-ID: <86f16dc10809010420p4df752eer9f487ad83bb552de@mail.gmail.com>

On Mon, Sep 1, 2008 at 3:13 AM, David Cournapeau <cournape at gmail.com> wrote:

> You should not have to care how it is stored, normally. audiolab gives
> you one column per channel; audiolab is just a wrapper around sndfile,
> which handles interleaving/deinterleaving internally if necessary.


I was referring to earlier, before I used audiolab and got an odd
spectrogram.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080901/33b1a710/attachment.html>

From cournape at gmail.com  Mon Sep  1 10:23:22 2008
From: cournape at gmail.com (David Cournapeau)
Date: Mon, 1 Sep 2008 23:23:22 +0900
Subject: [SciPy-user] install SciPy on Mac OS X 10.4
In-Reply-To: <8687ec0214b346df50502c9af4a8f9b4@85.18.140.153>
References: <8687ec0214b346df50502c9af4a8f9b4@85.18.140.153>
Message-ID: <5b8d13220809010723h47d34accgb9dedeade20348a4@mail.gmail.com>

On Mon, Sep 1, 2008 at 6:56 PM,  <luca.ciciriello at email.it> wrote:
>
> Hi All.
> I've installed NumPy 1.1.1 and now I want to install SciPy 0.6.0 on My Mac
> OS X
> (10.4 PPC). I've got scipy-0.6.0.tar.gz.tar from SciPy site. On My system
> is installed Xcode 2.5, so my default compiler is gcc 4.0.1. I've read in
> INSTALL.TXT in SciPy distribution that SciPy is not so compatible with this
> compiler, then to try to digit gcc_select 3.3.

Hm, those informations are really obsolete. Please ignore them, we
should update them.

cheers,

David


From c.j.lee at tnw.utwente.nl  Mon Sep  1 10:40:00 2008
From: c.j.lee at tnw.utwente.nl (Chris Lee)
Date: Mon, 1 Sep 2008 16:40:00 +0200
Subject: [SciPy-user] install SciPy on Mac OS X 10.4
In-Reply-To: <5b8d13220809010723h47d34accgb9dedeade20348a4@mail.gmail.com>
References: <8687ec0214b346df50502c9af4a8f9b4@85.18.140.153>
	<5b8d13220809010723h47d34accgb9dedeade20348a4@mail.gmail.com>
Message-ID: <F5E7CC8C-26B7-4279-A5F1-BB6250802F5D@tnw.utwente.nl>

The enthought distro will do all this for you as well.  (http://www.enthought.com/products/epdacademic.php?ver=MacOSX 
)

Cheers
Chris

On Sep 1, 2008, at 4:23 PM, David Cournapeau wrote:

> On Mon, Sep 1, 2008 at 6:56 PM,  <luca.ciciriello at email.it> wrote:
>>
>> Hi All.
>> I've installed NumPy 1.1.1 and now I want to install SciPy 0.6.0 on  
>> My Mac
>> OS X
>> (10.4 PPC). I've got scipy-0.6.0.tar.gz.tar from SciPy site. On My  
>> system
>> is installed Xcode 2.5, so my default compiler is gcc 4.0.1. I've  
>> read in
>> INSTALL.TXT in SciPy distribution that SciPy is not so compatible  
>> with this
>> compiler, then to try to digit gcc_select 3.3.
>
> Hm, those informations are really obsolete. Please ignore them, we
> should update them.
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

***************************************************
Chris Lee
Laser Physics and Nonlinear Optics Group
MESA+ Research Institute for Nanotechnology
University of Twente
Phone: ++31 (0)53 489 3968
fax: ++31 (0)53 489 1102
***************************************************





From jeff.lyon at cox.net  Mon Sep  1 11:41:52 2008
From: jeff.lyon at cox.net (Jeff Lyon)
Date: Mon, 1 Sep 2008 08:41:52 -0700
Subject: [SciPy-user] Create a spectrogram from a waveform
In-Reply-To: <20080830202812.bpnp4vvz0g4g4o80@mail.enthought.com>
References: <86f16dc10808301523x37dfe309y29d496564c6cb305@mail.gmail.com>
	<20080830202812.bpnp4vvz0g4g4o80@mail.enthought.com>
Message-ID: <C5589FE2-9CFB-48B1-8124-520E5CBB7CDC@cox.net>

Hello,

I tried to run the spectrogram.py example and I appear to be having  
configuration problems. I have installed the latest enthought distro,  
but the enable module can seem to find it's api component. Any thoughts?

~ jeff$ python
Enthought Python Distribution (2.5.2001) -- http://code.enthought.com

Python 2.5.2 |EPD 2.5.2001| (r252:60911, Jul  1 2008, 19:18:12)
[GCC 4.0.1 (Apple Computer, Inc. build 5370)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import spectrum.py
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
   File "spectrum.py", line 19, in <module>
     from enthought.enable.api import Window
ImportError: No module named api


Thanks,
Jeff Lyon


On Aug 30, 2008, at 6:28 PM, Peter Wang wrote:

> Quoting Ed McCaffrey <ed at edmccaffrey.net>:
>
>> I wrote a program in C# that creates a spectrogram from the  
>> waveform of a
>> .wav music file.  I now want to port it to Python, and I want to  
>> try to use
>> SciPy instead of a direct port of the existing code, because I am  
>> not sure
>> that it is perfectly accurate, and it is probably slow.
>>
>> I am having a hard time finding out how to do this with SciPy.   
>> With my
>> code, I had a FFT function that took an array of real and imaginary
>> components for each sample, and a second function taking both that  
>> produced
>> the amplitude.  The FFT function in SciPy just takes one array.
>>
>> Has anyone done this task in SciPy?
>
> We have a realtime spectrogram plot in the Audio Spectrum example for
> Chaco.  (See the very last screenshot on the gallery page here:
> http://code.enthought.com/projects/chaco/gallery.php)
>
> You can see the full source code of the example here:
> https://svn.enthought.com/enthought/browser/Chaco/trunk/examples/advanced/spectrum.py
>
> The lines you would be interested in are the last few:
>
> def get_audio_data():
>     pa = PyAudio()
>     stream = pa.open(format=paInt16, channels=1, rate=SAMPLING_RATE,
>                      input=True,
>                      frames_per_buffer=NUM_SAMPLES)
>     string_audio_data = stream.read(NUM_SAMPLES)
>     audio_data  = fromstring(string_audio_data, dtype=short)
>     normalized_data = audio_data / 32768.0
>     return (abs(fft(normalized_data))[:NUM_SAMPLES/2],  
> normalized_data)
>
> Here we are using the PyAudio library to directly read from the sound
> card, normalize the 16-bit data, and perform an FFT on it.
>
> In your case, since you are reading a WAV file, you might be
> interested in the zoomed_plot example:
> http://code.enthought.com/projects/chaco/pu-zooming-plot.html
>
> This displays the time-space signal but can easily be modified to show
> the FFT.  Here is the relevant code that uses the built-in python
> 'wave' module to read the data:
> https://svn.enthought.com/enthought/browser/Chaco/trunk/examples/zoomed_plot/wav_to_numeric.py
>
> You should be able to take the 'data' array in the wav_to_numeric
> function and hand that in to the fft function.
>
>
> -Peter
>
> _______________________________________________
> 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/20080901/77392e4d/attachment.html>

From bernardo.rocha at meduni-graz.at  Tue Sep  2 05:47:51 2008
From: bernardo.rocha at meduni-graz.at (bernardo martins rocha)
Date: Tue, 02 Sep 2008 11:47:51 +0200
Subject: [SciPy-user] matplotlib - axis and zoom
Message-ID: <48BD0BC7.9080502@meduni-graz.at>

Hi Guys,

how can I maintain zoom in the horizontal axis and keep the vertical 
axis unchanged using the zoom tool provided by matplotlib?

Bernardo M. Rocha


From cimrman3 at ntc.zcu.cz  Tue Sep  2 08:14:14 2008
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Tue, 02 Sep 2008 14:14:14 +0200
Subject: [SciPy-user] ANN: SfePy-00.50.00
Message-ID: <48BD2E16.4080307@ntc.zcu.cz>

I am pleased announce the release of SfePy 00.50.00.

SfePy is a finite element analysis software in Python, based primarily
on Numpy and SciPy.

Mailing lists, issue tracking, mercurial repository: http://sfepy.org
Home page: http://sfepy.kme.zcu.cz

People who contributed to this release: Ondrej Certik, Ryan Krauss, 
Vladimir Lukes.

Major improvements:
- finite strain elasticity: neo-Hookean, Mooney-Rivlin materials in the 
total Lagrangian (TL) formulation
- solving problems in complex numbers
- generalized equations to allow linear combination of terms
- run-time type of state term arguments
- refactoring to follow Python coding style guidelines
- new terms

For more information on this release, see
http://sfepy.googlecode.com/svn/web/releases/005000_RELEASE_NOTES.txt

Best regards,
Robert Cimrman


From christophe.grimault at novagrid.com  Tue Sep  2 13:11:22 2008
From: christophe.grimault at novagrid.com (christophe grimault)
Date: Tue, 02 Sep 2008 19:11:22 +0200
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <48BD2E16.4080307@ntc.zcu.cz>
References: <48BD2E16.4080307@ntc.zcu.cz>
Message-ID: <1220375482.3005.13.camel@pandora.novagrid.com>

Hi,

I have a application that is very demanding in memory ressources. So I
started to to look closer at python + numpy/scipy as far as memory is
concerned. 

I can't explain the following :

I start my python, + import scipy. A 'top' in the console shows that :

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME COMMAND
14791 grimault  20   0 21624 8044 3200 S    0  0.4   0:00.43 python

Now after typing :

z = scipy.arange(1000000)

I get :
14791 grimault  20   0 25532  11m 3204 S    0  0.6   0:00.44 python

So the memory increased by ~ 7 Mb. I was expecting 4 Mb since the data
type is int32, giving 4*1000000 = 4 Mb of memory chunk (in C/C++ at
least).

It gets even worse with complex float. I tried : 
z = arange(1000000) + 1j*arange(1000000)

Expecting 8 Mb, since z.dtype gives "complex64", the "top" shows an
increase by 31 Mb.

This is very annoying. Can someone explain this ? Is there a way to
create numpy arrays with the same (approximately ! I know the array
class adds some overhead...) memory footprint as in C/C++ ?

Thanks in advance






From robert.kern at gmail.com  Tue Sep  2 15:10:47 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 2 Sep 2008 14:10:47 -0500
Subject: [SciPy-user] matplotlib - axis and zoom
In-Reply-To: <48BD0BC7.9080502@meduni-graz.at>
References: <48BD0BC7.9080502@meduni-graz.at>
Message-ID: <3d375d730809021210p15f996ebq225228063f3a9450@mail.gmail.com>

On Tue, Sep 2, 2008 at 04:47, bernardo martins rocha
<bernardo.rocha at meduni-graz.at> wrote:
> Hi Guys,
>
> how can I maintain zoom in the horizontal axis and keep the vertical
> axis unchanged using the zoom tool provided by matplotlib?

You will want to ask on the matplotlib list:

  https://lists.sourceforge.net/lists/listinfo/matplotlib-users

-- 
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 cournape at gmail.com  Tue Sep  2 19:19:10 2008
From: cournape at gmail.com (David Cournapeau)
Date: Wed, 3 Sep 2008 08:19:10 +0900
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <1220375482.3005.13.camel@pandora.novagrid.com>
References: <48BD2E16.4080307@ntc.zcu.cz>
	<1220375482.3005.13.camel@pandora.novagrid.com>
Message-ID: <5b8d13220809021619g3927dc47y6bddb687e19e1ce8@mail.gmail.com>

On Wed, Sep 3, 2008 at 2:11 AM, christophe grimault
<christophe.grimault at novagrid.com> wrote:
> Hi,
>
> I have a application that is very demanding in memory ressources. So I
> started to to look closer at python + numpy/scipy as far as memory is
> concerned.

I you are really tight on memory, you will have problems with python
and most programming language which do not let you control memory in a
fine grained manner. Now, it depends on what you mean by memory
demanding: if you have barely enough memory for holding your data, it
will extremely difficult to do it in python, and difficult to do in
any language, including C and other manually managed languages.

>
> I can't explain the following :
>
> I start my python, + import scipy. A 'top' in the console shows that :
>
>  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME COMMAND
> 14791 grimault  20   0 21624 8044 3200 S    0  0.4   0:00.43 python
>
> Now after typing :
>
> z = scipy.arange(1000000)
>
> I get :
> 14791 grimault  20   0 25532  11m 3204 S    0  0.6   0:00.44 python
>
> So the memory increased by ~ 7 Mb. I was expecting 4 Mb since the data
> type is int32, giving 4*1000000 = 4 Mb of memory chunk (in C/C++ at
> least).

a = scipy.arange(1e6)
a.itemsize * a.size

Give me 8e6 bytes. arange is float64 by default, and I get a similar
memory increase (~ 8Mb).

>
> It gets even worse with complex float. I tried :
> z = arange(1000000) + 1j*arange(1000000)
>
> Expecting 8 Mb,

Again, this is strange, it should default to float128. Which version
of numpy/scipy are you using ?

I do not get unexpected results on my machine; results may vary
because memory allocator in python tends to overcommit to avoid
reallocating all the time, but IIRC, data are allocated with malloc
and not the python allocator in numpy. More importantly though, that's
not really representative of a typical numpy program, and would depend
on what you are doing anyway.

> This is very annoying. Can someone explain this ? Is there a way to
> create numpy arrays with the same (approximately ! I know the array
> class adds some overhead...) memory footprint as in C/C++ ?

Arrays themselves have a similar footprint as C/C++ (for big arrays,
where data >> array stucture overhead). But you will quickly find that
depending on what you are doing (linear algebra, for example), you
will need copies. Note that the same problem exists in C/C++, that's
very difficult to avoid (you need things like expression template and
co).

cheers,

David

cheers,

David


From robert.kern at gmail.com  Tue Sep  2 19:44:14 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 2 Sep 2008 18:44:14 -0500
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <5b8d13220809021619g3927dc47y6bddb687e19e1ce8@mail.gmail.com>
References: <48BD2E16.4080307@ntc.zcu.cz>
	<1220375482.3005.13.camel@pandora.novagrid.com>
	<5b8d13220809021619g3927dc47y6bddb687e19e1ce8@mail.gmail.com>
Message-ID: <3d375d730809021644k4232061ak327841959e0153a0@mail.gmail.com>

On Tue, Sep 2, 2008 at 18:19, David Cournapeau <cournape at gmail.com> wrote:
> On Wed, Sep 3, 2008 at 2:11 AM, christophe grimault
> <christophe.grimault at novagrid.com> wrote:
>> Hi,
>>
>> I have a application that is very demanding in memory ressources. So I
>> started to to look closer at python + numpy/scipy as far as memory is
>> concerned.
>
> I you are really tight on memory, you will have problems with python
> and most programming language which do not let you control memory in a
> fine grained manner. Now, it depends on what you mean by memory
> demanding: if you have barely enough memory for holding your data, it
> will extremely difficult to do it in python, and difficult to do in
> any language, including C and other manually managed languages.
>
>>
>> I can't explain the following :
>>
>> I start my python, + import scipy. A 'top' in the console shows that :
>>
>>  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME COMMAND
>> 14791 grimault  20   0 21624 8044 3200 S    0  0.4   0:00.43 python
>>
>> Now after typing :
>>
>> z = scipy.arange(1000000)
>>
>> I get :
>> 14791 grimault  20   0 25532  11m 3204 S    0  0.6   0:00.44 python
>>
>> So the memory increased by ~ 7 Mb. I was expecting 4 Mb since the data
>> type is int32, giving 4*1000000 = 4 Mb of memory chunk (in C/C++ at
>> least).
>
> a = scipy.arange(1e6)
> a.itemsize * a.size
>
> Give me 8e6 bytes. arange is float64 by default, and I get a similar
> memory increase (~ 8Mb).

No, the default is int (int32 on 32-bit systems, int64 on most 64-bit
systems) if you give it integer arguments and float64 if you give it
float arguments.

>> It gets even worse with complex float. I tried :
>> z = arange(1000000) + 1j*arange(1000000)
>>
>> Expecting 8 Mb,
>
> Again, this is strange, it should default to float128. Which version
> of numpy/scipy are you using ?

You mean complex128.

One thing to be aware of is that there are temporaries involved.
1j*arange(1000000) will allocate almost 16 MB of memory just by itself
and then allocate another 16 MB for the result of the addition. The
memory may not get returned to the OS when an object gets deallocated
although it will be reused by Python.

FWIW, here is what I get with SVN numpy on OS X:

>>> import numpy
45564 Python       0.0%  0:00.49   1    16    127 5172K  1292K  7960K    28M

>>> a = numpy.arange(1000000)
45564 Python       0.0%  0:00.50   1    16    128 9092K  1292K    12M    32M

>>> z = numpy.arange(1000000) + 1j * numpy.arange(1000000)
45564 Python       0.0%  0:00.60   1    16    129   24M  1292K    27M    47M

-- 
Robert Kern

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


From david at ar.media.kyoto-u.ac.jp  Tue Sep  2 21:19:22 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 03 Sep 2008 10:19:22 +0900
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <3d375d730809021644k4232061ak327841959e0153a0@mail.gmail.com>
References: <48BD2E16.4080307@ntc.zcu.cz>	<1220375482.3005.13.camel@pandora.novagrid.com>	<5b8d13220809021619g3927dc47y6bddb687e19e1ce8@mail.gmail.com>
	<3d375d730809021644k4232061ak327841959e0153a0@mail.gmail.com>
Message-ID: <48BDE61A.2000805@ar.media.kyoto-u.ac.jp>

Robert Kern wrote:
>
> No, the default is int (int32 on 32-bit systems, int64 on most 64-bit
> systems) if you give it integer arguments and float64 if you give it
> float arguments.
>   

Ah, my bad, I should have thought about the difference between 1e6 and
1000000.

>   
>>> It gets even worse with complex float. I tried :
>>> z = arange(1000000) + 1j*arange(1000000)
>>>
>>> Expecting 8 Mb,
>>>       
>> Again, this is strange, it should default to float128. Which version
>> of numpy/scipy are you using ?
>>     
>
> You mean complex128.
>   

Yes; I just wanted to point out that 1j*arange(1000000) is expected to
take ~16Mb, not 8.

> One thing to be aware of is that there are temporaries involved.
> 1j*arange(1000000) will allocate almost 16 MB of memory just by itself
> and then allocate another 16 MB for the result of the addition. The
> memory may not get returned to the OS when an object gets deallocated
> although it will be reused by Python.
>   

I think on linux, for those sizes, the memory is given back to the OS
right away because it is above the mmap threshold, and free gives the
memory right away in those cases. Since I see the exact same behavior as
you on top (b = np.arange(1e6) + 1.j np.arange(1e6) only adding 16 Mb),
maybe the Mac OS X malloc does something similar.

cheers,

David


From falted at pytables.org  Wed Sep  3 04:49:23 2008
From: falted at pytables.org (Francesc Alted)
Date: Wed, 3 Sep 2008 10:49:23 +0200
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <1220375482.3005.13.camel@pandora.novagrid.com>
References: <48BD2E16.4080307@ntc.zcu.cz>
	<1220375482.3005.13.camel@pandora.novagrid.com>
Message-ID: <200809031049.23777.falted@pytables.org>

A Tuesday 02 September 2008, christophe grimault escrigu?:
> Hi,
>
> I have a application that is very demanding in memory ressources. So
> I started to to look closer at python + numpy/scipy as far as memory
> is concerned.
>
> I can't explain the following :
>
> I start my python, + import scipy. A 'top' in the console shows that
> :
>
>   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME COMMAND
> 14791 grimault  20   0 21624 8044 3200 S    0  0.4   0:00.43 python
>
> Now after typing :
>
> z = scipy.arange(1000000)
>
> I get :
> 14791 grimault  20   0 25532  11m 3204 S    0  0.6   0:00.44 python
>
> So the memory increased by ~ 7 Mb. I was expecting 4 Mb since the
> data type is int32, giving 4*1000000 = 4 Mb of memory chunk (in C/C++
> at least).

You should look at the "RES" column instead of "VIRT" one.  "RES" column 
shows the *real* memory that you are consuming.  So, in this case, you 
have consumed 11MB - 8044KB ~ 3 MB.  However, you are undergoing here 
the effects of number representation truncation.  Your consumed memory 
should be rather: 8044KB + 3906KB = 11950KB, but as it is converted to 
MB (scale changes happens automatically in 'top' when the figures need 
more than 4 digits to be represented), 11959KB is truncated and 950KB 
are gone, so this is why the final figure you are seeing is 11MB.  This 
can be a bit misleading at first sight, but be sure that your machine 
(and NumPy) is doing fine and works as expected.

Cheers,

-- 
Francesc Alted


From christophe.grimault at novagrid.com  Wed Sep  3 05:49:50 2008
From: christophe.grimault at novagrid.com (christophe grimault)
Date: Wed, 03 Sep 2008 11:49:50 +0200
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <200809031049.23777.falted@pytables.org>
References: <48BD2E16.4080307@ntc.zcu.cz>
	<1220375482.3005.13.camel@pandora.novagrid.com>
	<200809031049.23777.falted@pytables.org>
Message-ID: <1220435390.2984.1.camel@pandora.novagrid.com>

OK, I understand now ! 
Thanks very for the explanation.

Chris

On Wed, 2008-09-03 at 10:49 +0200, Francesc Alted wrote:
> A Tuesday 02 September 2008, christophe grimault escrigu?:
> > Hi,
> >
> > I have a application that is very demanding in memory ressources. So
> > I started to to look closer at python + numpy/scipy as far as memory
> > is concerned.
> >
> > I can't explain the following :
> >
> > I start my python, + import scipy. A 'top' in the console shows that
> > :
> >
> >   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME COMMAND
> > 14791 grimault  20   0 21624 8044 3200 S    0  0.4   0:00.43 python
> >
> > Now after typing :
> >
> > z = scipy.arange(1000000)
> >
> > I get :
> > 14791 grimault  20   0 25532  11m 3204 S    0  0.6   0:00.44 python
> >
> > So the memory increased by ~ 7 Mb. I was expecting 4 Mb since the
> > data type is int32, giving 4*1000000 = 4 Mb of memory chunk (in C/C++
> > at least).
> 
> You should look at the "RES" column instead of "VIRT" one.  "RES" column 
> shows the *real* memory that you are consuming.  So, in this case, you 
> have consumed 11MB - 8044KB ~ 3 MB.  However, you are undergoing here 
> the effects of number representation truncation.  Your consumed memory 
> should be rather: 8044KB + 3906KB = 11950KB, but as it is converted to 
> MB (scale changes happens automatically in 'top' when the figures need 
> more than 4 digits to be represented), 11959KB is truncated and 950KB 
> are gone, so this is why the final figure you are seeing is 11MB.  This 
> can be a bit misleading at first sight, but be sure that your machine 
> (and NumPy) is doing fine and works as expected.
> 
> Cheers,
> 




From faltet at pytables.org  Wed Sep  3 06:36:44 2008
From: faltet at pytables.org (Francesc Alted)
Date: Wed, 3 Sep 2008 12:36:44 +0200
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <1220435390.2984.1.camel@pandora.novagrid.com>
References: <48BD2E16.4080307@ntc.zcu.cz>
	<200809031049.23777.falted@pytables.org>
	<1220435390.2984.1.camel@pandora.novagrid.com>
Message-ID: <200809031236.44488.faltet@pytables.org>

A Wednesday 03 September 2008, christophe grimault escrigu?:
> OK, I understand now !
> Thanks very for the explanation.

You are welcome.  For the sort of situations that you are facing of, I 
normally query directly the Linux kernel so as to get a finer 
perception of the resources wasted.  Here it goes the function that I 
use:

def show_stats(explain, tref):
    "Show the used memory"
    # Build the command to obtain memory info (only for Linux 2.6.x)
    cmd = "cat /proc/%s/status" % os.getpid()
    sout = subprocess.Popen(cmd, shell=True, 
stdout=subprocess.PIPE).stdout
    for line in sout:
        if line.startswith("VmSize:"):
            vmsize = int(line.split()[1])
        elif line.startswith("VmRSS:"):
            vmrss = int(line.split()[1])
        elif line.startswith("VmData:"):
            vmdata = int(line.split()[1])
        elif line.startswith("VmStk:"):
            vmstk = int(line.split()[1])
        elif line.startswith("VmExe:"):
            vmexe = int(line.split()[1])
        elif line.startswith("VmLib:"):
            vmlib = int(line.split()[1])
    sout.close()
    print "Memory usage: ******* %s *******" % explain
    print "VmSize: %7s kB\tVmRSS: %7s kB" % (vmsize, vmrss)
    print "VmData: %7s kB\tVmStk: %7s kB" % (vmdata, vmstk)
    print "VmExe:  %7s kB\tVmLib: %7s kB" % (vmexe, vmlib)
    tnow = time()
    print "WallClock time:", round(tnow - tref, 3)
    return tnow

And here it is an example of use:

# declare this at the beginning of your module
profile = True
[clip]
if profile: tref = time()
if profile: show_stats("Entering initial_append", tref)
[your statements here...]
if profile: show_stats("Before creating idx", tref)
[more statements...]
if profile: show_stats("After creating idx", tref)

I hope you get the idea.

-- 
Francesc Alted


From rocksportrocker at googlemail.com  Wed Sep  3 07:57:09 2008
From: rocksportrocker at googlemail.com (Uwe Schmitt)
Date: Wed, 3 Sep 2008 04:57:09 -0700 (PDT)
Subject: [SciPy-user] Arrays and strange memory usage ...
In-Reply-To: <1220375482.3005.13.camel@pandora.novagrid.com>
References: <48BD2E16.4080307@ntc.zcu.cz>
	<1220375482.3005.13.camel@pandora.novagrid.com>
Message-ID: <4da8b46e-00c6-4748-8c0d-fb035088e72a@m73g2000hsh.googlegroups.com>



On 2 Sep., 19:11, christophe grimault
<christophe.grima... at novagrid.com> wrote:
> Hi,
>
> I have a application that is very demanding in memory ressources. So I
> started to to look closer at python + numpy/scipy as far as memory is
> concerned.
>
> I can't explain the following :
>
> I start my python, + import scipy. A 'top' in the console shows that :
>
> ? PID USER ? ? ?PR ?NI ?VIRT ?RES ?SHR S %CPU %MEM ? ?TIME COMMAND
> 14791 grimault ?20 ? 0 21624 8044 3200 S ? ?0 ?0.4 ? 0:00.43 python
>
> Now after typing :
>
> z = scipy.arange(1000000)
>
> I get :
> 14791 grimault ?20 ? 0 25532 ?11m 3204 S ? ?0 ?0.6 ? 0:00.44 python
>
> So the memory increased by ~ 7 Mb. I was expecting 4 Mb since the data
> type is int32, giving 4*1000000 = 4 Mb of memory chunk (in C/C++ at
> least).

I do not see the 7MB. Virtual memory increased by 3,9 MB and RES
(which is the number you are looking for), differs by 3+X MB,
I do not know how RES is rounded. But this is not a contraction.

Greetings, Uwe


From emanuele at relativita.com  Fri Sep  5 11:23:24 2008
From: emanuele at relativita.com (Emanuele Olivetti)
Date: Fri, 05 Sep 2008 17:23:24 +0200
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
Message-ID: <48C14EEC.8060300@relativita.com>

Dear all and Dmitrey,

I've just updated to latest openopt (SVN). When using numpy 1.0.3
and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) openopt says
that "ralg" (NLP) algorithm is missing! With more recent numpy
and scipy it seems to work reliably. But what happened with respect
to older numpy+scipy? In that case even running examples/nlp_1.py
returns:
----
$ python nlp_1.py
OpenOpt checks user-supplied gradient df (shape: (150,) )
according to:
    prob.diffInt = [  1.00000000e-07]
    |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
according to:
    prob.diffInt = [  1.00000000e-07]
    |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
according to:
    prob.diffInt = [  1.00000000e-07]
    |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
derivatives are equal
========================
OO Error:incorrect solver is called, maybe the solver "ralg" is not
installed. Maybe setting p.debug=1 could specify the matter more precisely
Traceback (most recent call last):
  File "nlp_1.py", line 110, in <module>
    r = p.solve('ralg')
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
line 185, in solve
    return runProbSolver(self, solvers, *args, **kwargs)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
line 48, in runProbSolver
    p.err('incorrect solver is called, maybe the solver "' + solver_str
+'" is not installed. Maybe setting p.debug=1 could specify the matter
more precisely')
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
line 16, in ooerr
    raise OpenOptException(msg)
scikits.openopt.Kernel.oologfcn.OpenOptException: incorrect solver is
called, maybe the solver "ralg" is not installed. Maybe setting
p.debug=1 could specify the matter more precisely
----

This did not happen before so I guess it is due to a recent
commit. It is possible to solve the problem?

Kind Regards,

Emanuele



From cohen at slac.stanford.edu  Fri Sep  5 11:36:53 2008
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Fri, 05 Sep 2008 17:36:53 +0200
Subject: [SciPy-user] how to plot the result of histogram2d
Message-ID: <48C15215.3060605@slac.stanford.edu>

hi, I hope someone can quickly point me to some doc.
I can do imshow(histogram2d(x,y)[0]) but then I miss the x and y binning 
correct labels.
If I do imshow(histogram2d(x,y)) I get:
ERROR: An unexpected error occurred while tokenizing input
The following traceback may be corrupted or invalid
The error message is: ('EOF in multi-line statement', (115, 0))

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)

/home/cohen/data1/WORK/GRB/data/GRB080904886/<ipython console> in <module>()

/data1/GLAST/GLAST_EXT/python/2.5.1/lib/python2.5/site-packages/matplotlib/pyplot.pyc 
in imshow(*args, **kwargs)
   1673         hold(h)
   1674     try:
-> 1675         ret =  gca().imshow(*args, **kwargs)
   1676         draw_if_interactive()
   1677     except:

/data1/GLAST/GLAST_EXT/python/2.5.1/lib/python2.5/site-packages/matplotlib/axes.pyc 
in imshow(self, X, cmap, norm, aspect, interpolation, alpha, vmin, vmax, 
origin, extent, shape, filternorm, filterrad, imlim, **kwargs)
   4432                        filterrad=filterrad, **kwargs)
   4433
-> 4434         im.set_data(X)
   4435         im.set_alpha(alpha)
   4436         self._set_artist_props(im)

/data1/GLAST/GLAST_EXT/python/2.5.1/lib/python2.5/site-packages/matplotlib/image.pyc 
in set_data(self, A, shape)
    232             X = pil_to_array(A)
    233         else:
--> 234             X = ma.asarray(A) # assume array
    235         self._A = X
    236

/data1/GLAST/GLAST_EXT/python/2.5.1/lib/python2.5/site-packages/numpy/core/ma.pyc 
in asarray(data, dtype)
   2121         (dtype is None or dtype == data.dtype):
   2122         return data
-> 2123     return array(data, dtype=dtype, copy=0)
   2124
   2125 # Add methods to support ndarray interface

/data1/GLAST/GLAST_EXT/python/2.5.1/lib/python2.5/site-packages/numpy/core/ma.pyc 
in __init__(self, data, dtype, copy, order, mask, fill_value)
    565         else:
    566             need_data_copied = False #because I'll do it now
--> 567             c = numeric.array(data, dtype=tc, copy=True, 
order=order)
    568             tc = c.dtype
    569

ValueError: setting an array element with a sequence.


so something gets awry.
thanks in advance


From jdh2358 at gmail.com  Fri Sep  5 11:49:32 2008
From: jdh2358 at gmail.com (John Hunter)
Date: Fri, 5 Sep 2008 10:49:32 -0500
Subject: [SciPy-user] how to plot the result of histogram2d
In-Reply-To: <48C15215.3060605@slac.stanford.edu>
References: <48C15215.3060605@slac.stanford.edu>
Message-ID: <88e473830809050849n1bbc6beawde24bc19ad3b038c@mail.gmail.com>

On Fri, Sep 5, 2008 at 10:36 AM, Johann Cohen-Tanugi
<cohen at slac.stanford.edu> wrote:
> hi, I hope someone can quickly point me to some doc.
> I can do imshow(histogram2d(x,y)[0]) but then I miss the x and y binning
> correct labels.
> If I do imshow(histogram2d(x,y)) I get:
> ERROR: An unexpected error occurred while tokenizing input
> The following traceback may be corrupted or invalid
> The error message is: ('EOF in multi-line statement', (115, 0))

matplotlib questions are best addressed to the matplotlib-users mailing list at

   http://lists.sourceforge.net/mailman/listinfo/matplotlib-users

histogram2d returns H, xedges and yedges.  The first argument should
be passed to imshow, and the second two can be used to get the extents

In [26]: x, y = np.random.randn(2, 100000)

In [27]: H, xedges, yedges = np.histogram2d(x, y, bins=50)

In [28]: extent = [xedges[0], xedges[-1], yedges[0], yedges[-1]]

In [29]: imshow(H, extent=extent)
Out[29]: <matplotlib.image.AxesImage object at 0x9377bcc>

I


From emanuele at relativita.com  Fri Sep  5 11:52:31 2008
From: emanuele at relativita.com (Emanuele Olivetti)
Date: Fri, 05 Sep 2008 17:52:31 +0200
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C14EEC.8060300@relativita.com>
References: <48C14EEC.8060300@relativita.com>
Message-ID: <48C155BF.7080008@relativita.com>

Same problem with numpy 1.0.4 + scipy 0.6.0
(shipped with ubuntu 8.04 hardy heron).

E.

Emanuele Olivetti wrote:
> Dear all and Dmitrey,
>
> I've just updated to latest openopt (SVN). When using numpy 1.0.3
> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) openopt says
> that "ralg" (NLP) algorithm is missing! With more recent numpy
> and scipy it seems to work reliably. But what happened with respect
> to older numpy+scipy? In that case even running examples/nlp_1.py
> returns:
> ----
> $ python nlp_1.py
> OpenOpt checks user-supplied gradient df (shape: (150,) )
> according to:
>     prob.diffInt = [  1.00000000e-07]
>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
> derivatives are equal
> ========================
> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
> according to:
>     prob.diffInt = [  1.00000000e-07]
>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
> derivatives are equal
> ========================
> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
> according to:
>     prob.diffInt = [  1.00000000e-07]
>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
> derivatives are equal
> ========================
> OO Error:incorrect solver is called, maybe the solver "ralg" is not
> installed. Maybe setting p.debug=1 could specify the matter more precisely
> Traceback (most recent call last):
>   File "nlp_1.py", line 110, in <module>
>     r = p.solve('ralg')
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
> line 185, in solve
>     return runProbSolver(self, solvers, *args, **kwargs)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
> line 48, in runProbSolver
>     p.err('incorrect solver is called, maybe the solver "' + solver_str
> +'" is not installed. Maybe setting p.debug=1 could specify the matter
> more precisely')
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
> line 16, in ooerr
>     raise OpenOptException(msg)
> scikits.openopt.Kernel.oologfcn.OpenOptException: incorrect solver is
> called, maybe the solver "ralg" is not installed. Maybe setting
> p.debug=1 could specify the matter more precisely
> ----
>
> This did not happen before so I guess it is due to a recent
> commit. It is possible to solve the problem?
>
> Kind Regards,
>
> Emanuele
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   



From dmitrey.kroshko at scipy.org  Fri Sep  5 13:28:50 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Fri, 05 Sep 2008 20:28:50 +0300
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C155BF.7080008@relativita.com>
References: <48C14EEC.8060300@relativita.com> <48C155BF.7080008@relativita.com>
Message-ID: <48C16C52.1040101@scipy.org>

Hi Emanuele,
 as it is mentioned in openopt install webpage and README.txt  numpy v 
 >= 1.1.0 is recommended. Some other oo users informed of bugs due to 
older versions.

Could you inform what will be outputed if you set p.debug = 1? (either 
directly or via p = NLP(..., debug=1,...))

If the problem with numpy versions is critical for users of your soft, 
you'd better to put more recent numpy into Debian soft channel.

Regards, D.

Emanuele Olivetti wrote:
> Same problem with numpy 1.0.4 + scipy 0.6.0
> (shipped with ubuntu 8.04 hardy heron).
>
> E.
>
> Emanuele Olivetti wrote:
>   
>> Dear all and Dmitrey,
>>
>> I've just updated to latest openopt (SVN). When using numpy 1.0.3
>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) openopt says
>> that "ralg" (NLP) algorithm is missing! With more recent numpy
>> and scipy it seems to work reliably. But what happened with respect
>> to older numpy+scipy? In that case even running examples/nlp_1.py
>> returns:
>> ----
>> $ python nlp_1.py
>> OpenOpt checks user-supplied gradient df (shape: (150,) )
>> according to:
>>     prob.diffInt = [  1.00000000e-07]
>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>> derivatives are equal
>> ========================
>> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
>> according to:
>>     prob.diffInt = [  1.00000000e-07]
>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>> derivatives are equal
>> ========================
>> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
>> according to:
>>     prob.diffInt = [  1.00000000e-07]
>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>> derivatives are equal
>> ========================
>> OO Error:incorrect solver is called, maybe the solver "ralg" is not
>> installed. Maybe setting p.debug=1 could specify the matter more precisely
>> Traceback (most recent call last):
>>   File "nlp_1.py", line 110, in <module>
>>     r = p.solve('ralg')
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>> line 185, in solve
>>     return runProbSolver(self, solvers, *args, **kwargs)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>> line 48, in runProbSolver
>>     p.err('incorrect solver is called, maybe the solver "' + solver_str
>> +'" is not installed. Maybe setting p.debug=1 could specify the matter
>> more precisely')
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>> line 16, in ooerr
>>     raise OpenOptException(msg)
>> scikits.openopt.Kernel.oologfcn.OpenOptException: incorrect solver is
>> called, maybe the solver "ralg" is not installed. Maybe setting
>> p.debug=1 could specify the matter more precisely
>> ----
>>
>> This did not happen before so I guess it is due to a recent
>> commit. It is possible to solve the problem?
>>
>> Kind Regards,
>>
>> Emanuele
>>
>> _______________________________________________
>> 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 nwagner at iam.uni-stuttgart.de  Fri Sep  5 14:12:57 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 05 Sep 2008 20:12:57 +0200
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C16C52.1040101@scipy.org>
References: <48C14EEC.8060300@relativita.com> <48C155BF.7080008@relativita.com>
	<48C16C52.1040101@scipy.org>
Message-ID: <web-110798387@uni-stuttgart.de>

On Fri, 05 Sep 2008 20:28:50 +0300
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> Hi Emanuele,
> as it is mentioned in openopt install webpage and 
>README.txt  numpy v 
> >= 1.1.0 is recommended. Some other oo users informed of 
>bugs due to 
> older versions.
> 
> Could you inform what will be outputed if you set 
>p.debug = 1? (either 
> directly or via p = NLP(..., debug=1,...))
> 
> If the problem with numpy versions is critical for users 
>of your soft, 
> you'd better to put more recent numpy into Debian soft 
>channel.
> 
> Regards, D.
> 
> Emanuele Olivetti wrote:
>> Same problem with numpy 1.0.4 + scipy 0.6.0
>> (shipped with ubuntu 8.04 hardy heron).
>>
>> E.
>>
>> Emanuele Olivetti wrote:
>>   
>>> Dear all and Dmitrey,
>>>
>>> I've just updated to latest openopt (SVN). When using 
>>>numpy 1.0.3
>>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) 
>>>openopt says
>>> that "ralg" (NLP) algorithm is missing! With more recent 
>>>numpy
>>> and scipy it seems to work reliably. But what happened 
>>>with respect
>>> to older numpy+scipy? In that case even running 
>>>examples/nlp_1.py
>>> returns:
>>> ----
>>> $ python nlp_1.py
>>> OpenOpt checks user-supplied gradient df (shape: (150,) 
>>>)
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>= 0.01
>>> derivatives are equal
>>> ========================
>>> OpenOpt checks user-supplied gradient dc (shape: (2, 
>>>150) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>= 0.01
>>> derivatives are equal
>>> ========================
>>> OpenOpt checks user-supplied gradient dh (shape: (2, 
>>>150) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>= 0.01
>>> derivatives are equal
>>> ========================
>>> OO Error:incorrect solver is called, maybe the solver 
>>>"ralg" is not
>>> installed. Maybe setting p.debug=1 could specify the 
>>>matter more precisely
>>> Traceback (most recent call last):
>>>   File "nlp_1.py", line 110, in <module>
>>>     r = p.solve('ralg')
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>> line 185, in solve
>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>> line 48, in runProbSolver
>>>     p.err('incorrect solver is called, maybe the solver 
>>>"' + solver_str
>>> +'" is not installed. Maybe setting p.debug=1 could 
>>>specify the matter
>>> more precisely')
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>>> line 16, in ooerr
>>>     raise OpenOptException(msg)
>>> scikits.openopt.Kernel.oologfcn.OpenOptException: 
>>>incorrect solver is
>>> called, maybe the solver "ralg" is not installed. Maybe 
>>>setting
>>> p.debug=1 could specify the matter more precisely
>>> ----
>>>
>>> This did not happen before so I guess it is due to a 
>>>recent
>>> commit. It is possible to solve the problem?
>>>
>>> Kind Regards,
>>>
>>> Emanuele
>>>
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>
>>>   
>>>     
>>
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>>
>>
>>   
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

  
Dmitrey,

I am using
>>> numpy.__version__
'1.3.0.dev5790'

Cheers,
          Nils

Here comes the output of nlp_1.py:

OpenOpt checks user-supplied gradient df (shape: (150,) )
according to:
     prob.diffInt = [  1.00000000e-07]
     |1 - info_user/info_numerical| <= prob.maxViolation = 
0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dc (shape: (2, 150) 
)
according to:
     prob.diffInt = [  1.00000000e-07]
     |1 - info_user/info_numerical| <= prob.maxViolation = 
0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dh (shape: (2, 150) 
)
according to:
     prob.diffInt = [  1.00000000e-07]
     |1 - info_user/info_numerical| <= prob.maxViolation = 
0.01
derivatives are equal
========================
-----------------------------------------------------
solver: ralg   problem: unnamed   goal: minimum
  iter    objFunVal    log10(maxResidual)
     0  8.596e+03               3.91
OpenOpt debug msg:  hs: 4.0
OpenOpt debug msg:  ls: 2
    50  2.800e+03               0.79
   100  1.754e+03               0.52
   150  9.075e+02               0.31
   200  4.455e+02              -0.03
   250  3.682e+02              -0.48
   300  3.465e+02              -1.15
   350  3.409e+02              -1.81
   400  1.911e+02              -3.14
   450  1.373e+02              -3.07
OO info:  debug msg: matrix B restoration in ralg solver
   500  1.065e+03               1.20
   550  2.224e+03               1.21
   600  1.822e+03               0.43
   650  2.178e+03               0.45
   700  2.576e+03               0.48
   750  2.840e+03               0.53
   800  3.068e+03               0.59
   850  7.958e+03               1.37
   900  2.174e+04               1.54
   950  3.341e+04               1.37
  1000  7.463e+04               2.17
  1050  3.692e+05               2.50
  1100  1.940e+05               2.16
  1150  1.482e+05               1.77
  1200  1.719e+05               1.86
  1250  2.963e+05               2.52
  1300  1.603e+05               2.27
  1350  2.299e+05               2.56
  1400  3.243e+05               2.63
  1450  2.663e+05               2.51
  1500  3.064e+05               2.55
  1550  4.297e+05               2.74
  1600  1.629e+05               2.80
  1650  2.379e+05               2.33
  1700  2.086e+05               2.28
  1750  1.214e+05               2.22
  1800  4.913e+04               1.58
  1850  3.862e+04               1.65
  1900  1.610e+05               2.53
  1950  3.576e+04               1.44
OO info:  debug msg: matrix B restoration in ralg solver
  2000  7.286e+05               2.42
  2050  5.268e+05               2.50
  2100  1.403e+05               2.01
  2150  1.029e+05               1.96
  2200  9.997e+04               2.15
  2250  7.424e+05               2.92
  2300  5.514e+04               1.55
  2350  2.518e+05               2.66
  2400  5.051e+04               1.78
  2450  5.006e+04               2.05
  2500  4.279e+04               1.44
  2550  4.509e+04               1.62
  2600  1.331e+05               2.45
  2650  4.061e+04               1.41
  2700  5.198e+04               1.90
  2750  3.489e+09               4.77
  2800  6.938e+04               2.22
  2850  2.474e+10               5.20
  2900  4.259e+07               3.81
  2950  1.044e+05               2.40
  3000  6.411e+10               5.40
  3050  6.232e+07               3.89
  3100  1.830e+06               3.13
  3150  4.635e+04               1.45
  3200  1.770e+09               4.62
OO info:  debug msg: matrix B restoration in ralg solver
  3250  1.764e+11               5.57
  3300  3.792e+09               4.01
  3350  1.554e+10               5.05
  3400  7.294e+09               4.81
  3450  7.227e+09               4.81
OO info:  debug msg: matrix B restoration in ralg solver
  3500  1.415e+11               5.56
  3550  1.795e+10               6.16
  3600  5.205e+09               4.40
  3650  1.641e+10               5.04
  3700  1.408e+10               5.01
OO info:  debug msg: matrix B restoration in ralg solver
  3750  1.277e+10               4.96
  3800  5.576e+09               3.89
  3850  5.008e+09               3.97
  3900  4.475e+09               4.04
  3950  3.993e+09               4.11
  4000  3.558e+09               4.17
  4050  3.237e+09               4.24
  4100  2.844e+09               4.24
  4150  1.077e+10               4.83
  4200  9.891e+09               4.82
OO info:  debug msg: matrix B restoration in ralg solver
  4250  4.720e+09               4.12
  4300  3.411e+09               4.02
  4350  1.768e+09               6.43
  4400  1.851e+09               4.31
  4450  1.448e+09               3.99
  4500  1.248e+09               3.84
  4550  1.099e+09               3.80
  4600  6.053e+09               4.85
  4650  8.905e+08               3.86
  4700  1.446e+09               4.43
OO info:  debug msg: matrix B restoration in ralg solver
  4750  6.292e+09               4.14
  4800  2.558e+09               3.96
  4850  2.898e+09               4.53
  4900  1.581e+09               4.21
  4950  1.272e+09               4.28
  5000  5.860e+09               6.34
  5050  4.163e+09               4.56
  5100  3.478e+09               4.22
  5150  3.238e+09               4.31
  5200  2.862e+09               3.92
OO info:  debug msg: matrix B restoration in ralg solver
  5250  3.259e+09               4.36
  5300  2.207e+09               3.91
  5350  1.760e+09               3.74
  5400  1.560e+09               3.93
  5450  1.925e+09               4.41
  5500  1.739e+09               4.41
  5550  1.640e+09               4.42
  5600  8.408e+10               4.93
  5650  9.792e+10               4.69
  5700  1.303e+11               4.75
  5750  2.450e+11               5.44
  5800  4.913e+11               5.33
  5850  2.536e+11               6.00
  5900  3.098e+11               5.70
  5950  8.987e+10               5.37
OO info:  debug msg: matrix B restoration in ralg solver
  6000  1.037e+12               6.00
  6050  3.448e+11               8.99
  6100  8.307e+12               6.40
  6150  1.589e+12               5.87
  6200  1.213e+12               5.27
OO info:  debug msg: matrix B restoration in ralg solver
  6250  1.224e+12               5.45
  6300  7.495e+11               5.00
  6350  3.998e+11              15.67
  6400  3.987e+11               5.00
  6450  3.127e+11               5.02
  6500  2.419e+11               5.27
  6550  3.691e+11               5.13
  6600  6.414e+11               5.74
  6650  1.329e+12               5.92
  6700  3.528e+11               5.18
  6750  2.981e+11               4.78
  6800  5.060e+11               5.51
  6850  4.760e+11               5.09
  6900  4.499e+11               5.10
  6950  1.069e+12               5.86
  7000  6.326e+11               5.26
  7050  5.217e+11               5.18
  7100  5.029e+11               5.16
  7150  8.043e+12               6.43
  7200  1.073e+13               6.51
  7250  2.658e+12               6.18
  7300  2.053e+11               4.81
  7350  1.040e+12               5.45
  7400  2.030e+12               6.08
  7450  2.131e+12               6.11
  7500  3.493e+11               5.17
  7550  2.420e+11               5.04
  7600  2.344e+12               6.17
  7650  3.515e+11               5.62
  7700  2.135e+11               5.35
  7750  1.411e+11               4.78
  7800  8.295e+12               6.46
  7850  7.406e+12               6.39
  7900  9.030e+12               6.45
  7950  1.677e+12               6.04
OO info:  debug msg: matrix B restoration in ralg solver
  8000  3.579e+12               6.23
  8050  1.109e+12              10.92
  8100  5.111e+12               5.80
  8150  7.521e+12               6.08
  8200  7.199e+12               5.85
OO info:  debug msg: matrix B restoration in ralg solver
  8250  7.812e+12               6.05
  8300  5.366e+12               8.57
  8350  5.689e+12               5.97
  8400  5.140e+12               5.97
  8450  3.909e+12               5.38
OO info:  debug msg: matrix B restoration in ralg solver
  8500  5.130e+12               6.12
  8550  3.753e+12               6.36
  8600  2.963e+12               5.43
  8650  2.528e+12               5.44
  8700  2.134e+12               5.46
OO info:  debug msg: matrix B restoration in ralg solver
  8750  1.760e+12               5.46
  8800  1.467e+12               5.27
  8850  2.764e+12              12.53
  8900  2.152e+12               5.63
  8950  2.532e+12               5.86
OO info:  debug msg: matrix B restoration in ralg solver
  9000  1.884e+12               5.67
  9050  4.073e+12              12.35
  9100  1.709e+12               5.38
  9150  1.398e+12               5.57
  9200  1.248e+12               5.60
OO info:  debug msg: matrix B restoration in ralg solver
  9250  1.044e+12               5.14
  9300  7.844e+11               5.21
  9350  6.360e+11               5.47
  9400  6.253e+11               5.67
  9450  3.557e+11               4.91
  9500  3.400e+11               5.29
  9550  3.160e+11               5.30
  9600  2.601e+11               4.94
  9650  2.199e+11               4.85
  9700  5.335e+12              13.48
OO info:  debug msg: matrix B restoration in ralg solver
  9750  5.933e+12               6.24
  9800  4.174e+12               8.76
  9850  3.803e+12               5.52
  9900  2.854e+12               5.50
  9950  2.014e+12               5.47
10000  3.285e+12               6.13
10001  3.285e+12               6.13
istop:  -7 (Max Iter has been reached)
Solver:   Time Elapsed = 56.05  CPU Time Elapsed = 31.82
Plotting: Time Elapsed = 62.35  CPU Time Elapsed = 32.57
NO FEASIBLE SOLUTION is obtained (max residual = 1.4e+06, 
objFunc = 3.2852899e+12)


From dmitrey.kroshko at scipy.org  Fri Sep  5 15:40:12 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Fri, 05 Sep 2008 22:40:12 +0300
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <web-110798387@uni-stuttgart.de>
References: <48C14EEC.8060300@relativita.com>
	<48C155BF.7080008@relativita.com>	<48C16C52.1040101@scipy.org>
	<web-110798387@uni-stuttgart.de>
Message-ID: <48C18B1C.4030801@scipy.org>

Hi Nils,
after some my modifications the file nlp_1.py become hard to be solved 
by any solver (I mean connected to OO) - you can try solving it by 
algencan or ipopt and see results (output).

So I have committed some changes to nlp_1.py. As for other tests (like 
nlp_bench_1, nlp_3) they work ok (nlp_2 for ralg requires p.maxIter = 2000).

Regards, D.

Nils Wagner wrote:
> On Fri, 05 Sep 2008 20:28:50 +0300
>   dmitrey <dmitrey.kroshko at scipy.org> wrote:
>   
>> Hi Emanuele,
>> as it is mentioned in openopt install webpage and 
>> README.txt  numpy v 
>>     
>>> = 1.1.0 is recommended. Some other oo users informed of 
>>>       
>> bugs due to 
>> older versions.
>>
>> Could you inform what will be outputed if you set 
>> p.debug = 1? (either 
>> directly or via p = NLP(..., debug=1,...))
>>
>> If the problem with numpy versions is critical for users 
>> of your soft, 
>> you'd better to put more recent numpy into Debian soft 
>> channel.
>>
>> Regards, D.
>>
>> Emanuele Olivetti wrote:
>>     
>>> Same problem with numpy 1.0.4 + scipy 0.6.0
>>> (shipped with ubuntu 8.04 hardy heron).
>>>
>>> E.
>>>
>>> Emanuele Olivetti wrote:
>>>   
>>>       
>>>> Dear all and Dmitrey,
>>>>
>>>> I've just updated to latest openopt (SVN). When using 
>>>> numpy 1.0.3
>>>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) 
>>>> openopt says
>>>> that "ralg" (NLP) algorithm is missing! With more recent 
>>>> numpy
>>>> and scipy it seems to work reliably. But what happened 
>>>> with respect
>>>> to older numpy+scipy? In that case even running 
>>>> examples/nlp_1.py
>>>> returns:
>>>> ----
>>>> $ python nlp_1.py
>>>> OpenOpt checks user-supplied gradient df (shape: (150,) 
>>>> )
>>>> according to:
>>>>     prob.diffInt = [  1.00000000e-07]
>>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>> = 0.01
>>>> derivatives are equal
>>>> ========================
>>>> OpenOpt checks user-supplied gradient dc (shape: (2, 
>>>> 150) )
>>>> according to:
>>>>     prob.diffInt = [  1.00000000e-07]
>>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>> = 0.01
>>>> derivatives are equal
>>>> ========================
>>>> OpenOpt checks user-supplied gradient dh (shape: (2, 
>>>> 150) )
>>>> according to:
>>>>     prob.diffInt = [  1.00000000e-07]
>>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>> = 0.01
>>>> derivatives are equal
>>>> ========================
>>>> OO Error:incorrect solver is called, maybe the solver 
>>>> "ralg" is not
>>>> installed. Maybe setting p.debug=1 could specify the 
>>>> matter more precisely
>>>> Traceback (most recent call last):
>>>>   File "nlp_1.py", line 110, in <module>
>>>>     r = p.solve('ralg')
>>>>   File
>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>>> line 185, in solve
>>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>>   File
>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>>> line 48, in runProbSolver
>>>>     p.err('incorrect solver is called, maybe the solver 
>>>> "' + solver_str
>>>> +'" is not installed. Maybe setting p.debug=1 could 
>>>> specify the matter
>>>> more precisely')
>>>>   File
>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>>>> line 16, in ooerr
>>>>     raise OpenOptException(msg)
>>>> scikits.openopt.Kernel.oologfcn.OpenOptException: 
>>>> incorrect solver is
>>>> called, maybe the solver "ralg" is not installed. Maybe 
>>>> setting
>>>> p.debug=1 could specify the matter more precisely
>>>> ----
>>>>
>>>> This did not happen before so I guess it is due to a 
>>>> recent
>>>> commit. It is possible to solve the problem?
>>>>
>>>> Kind Regards,
>>>>
>>>> Emanuele
>>>>
>>>> _______________________________________________
>>>> SciPy-user mailing list
>>>> SciPy-user at scipy.org
>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>
>>>>   
>>>>     
>>>>         
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>
>>>
>>>
>>>   
>>>       
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>     
>
>   
> Dmitrey,
>
> I am using
>   
>>>> numpy.__version__
>>>>         
> '1.3.0.dev5790'
>
> Cheers,
>           Nils
>
> Here comes the output of nlp_1.py:
>
> OpenOpt checks user-supplied gradient df (shape: (150,) )
> according to:
>      prob.diffInt = [  1.00000000e-07]
>      |1 - info_user/info_numerical| <= prob.maxViolation = 
> 0.01
> derivatives are equal
> ========================
> OpenOpt checks user-supplied gradient dc (shape: (2, 150) 
> )
> according to:
>      prob.diffInt = [  1.00000000e-07]
>      |1 - info_user/info_numerical| <= prob.maxViolation = 
> 0.01
> derivatives are equal
> ========================
> OpenOpt checks user-supplied gradient dh (shape: (2, 150) 
> )
> according to:
>      prob.diffInt = [  1.00000000e-07]
>      |1 - info_user/info_numerical| <= prob.maxViolation = 
> 0.01
> derivatives are equal
> ========================
> -----------------------------------------------------
> solver: ralg   problem: unnamed   goal: minimum
>   iter    objFunVal    log10(maxResidual)
>      0  8.596e+03               3.91
> OpenOpt debug msg:  hs: 4.0
> OpenOpt debug msg:  ls: 2
>     50  2.800e+03               0.79
>    100  1.754e+03               0.52
>    150  9.075e+02               0.31
>    200  4.455e+02              -0.03
>    250  3.682e+02              -0.48
>    300  3.465e+02              -1.15
>    350  3.409e+02              -1.81
>    400  1.911e+02              -3.14
>    450  1.373e+02              -3.07
> OO info:  debug msg: matrix B restoration in ralg solver
>    500  1.065e+03               1.20
>    550  2.224e+03               1.21
>    600  1.822e+03               0.43
>    650  2.178e+03               0.45
>    700  2.576e+03               0.48
>    750  2.840e+03               0.53
>    800  3.068e+03               0.59
>    850  7.958e+03               1.37
>    900  2.174e+04               1.54
>    950  3.341e+04               1.37
>   1000  7.463e+04               2.17
>   1050  3.692e+05               2.50
>   1100  1.940e+05               2.16
>   1150  1.482e+05               1.77
>   1200  1.719e+05               1.86
>   1250  2.963e+05               2.52
>   1300  1.603e+05               2.27
>   1350  2.299e+05               2.56
>   1400  3.243e+05               2.63
>   1450  2.663e+05               2.51
>   1500  3.064e+05               2.55
>   1550  4.297e+05               2.74
>   1600  1.629e+05               2.80
>   1650  2.379e+05               2.33
>   1700  2.086e+05               2.28
>   1750  1.214e+05               2.22
>   1800  4.913e+04               1.58
>   1850  3.862e+04               1.65
>   1900  1.610e+05               2.53
>   1950  3.576e+04               1.44
> OO info:  debug msg: matrix B restoration in ralg solver
>   2000  7.286e+05               2.42
>   2050  5.268e+05               2.50
>   2100  1.403e+05               2.01
>   2150  1.029e+05               1.96
>   2200  9.997e+04               2.15
>   2250  7.424e+05               2.92
>   2300  5.514e+04               1.55
>   2350  2.518e+05               2.66
>   2400  5.051e+04               1.78
>   2450  5.006e+04               2.05
>   2500  4.279e+04               1.44
>   2550  4.509e+04               1.62
>   2600  1.331e+05               2.45
>   2650  4.061e+04               1.41
>   2700  5.198e+04               1.90
>   2750  3.489e+09               4.77
>   2800  6.938e+04               2.22
>   2850  2.474e+10               5.20
>   2900  4.259e+07               3.81
>   2950  1.044e+05               2.40
>   3000  6.411e+10               5.40
>   3050  6.232e+07               3.89
>   3100  1.830e+06               3.13
>   3150  4.635e+04               1.45
>   3200  1.770e+09               4.62
> OO info:  debug msg: matrix B restoration in ralg solver
>   3250  1.764e+11               5.57
>   3300  3.792e+09               4.01
>   3350  1.554e+10               5.05
>   3400  7.294e+09               4.81
>   3450  7.227e+09               4.81
> OO info:  debug msg: matrix B restoration in ralg solver
>   3500  1.415e+11               5.56
>   3550  1.795e+10               6.16
>   3600  5.205e+09               4.40
>   3650  1.641e+10               5.04
>   3700  1.408e+10               5.01
> OO info:  debug msg: matrix B restoration in ralg solver
>   3750  1.277e+10               4.96
>   3800  5.576e+09               3.89
>   3850  5.008e+09               3.97
>   3900  4.475e+09               4.04
>   3950  3.993e+09               4.11
>   4000  3.558e+09               4.17
>   4050  3.237e+09               4.24
>   4100  2.844e+09               4.24
>   4150  1.077e+10               4.83
>   4200  9.891e+09               4.82
> OO info:  debug msg: matrix B restoration in ralg solver
>   4250  4.720e+09               4.12
>   4300  3.411e+09               4.02
>   4350  1.768e+09               6.43
>   4400  1.851e+09               4.31
>   4450  1.448e+09               3.99
>   4500  1.248e+09               3.84
>   4550  1.099e+09               3.80
>   4600  6.053e+09               4.85
>   4650  8.905e+08               3.86
>   4700  1.446e+09               4.43
> OO info:  debug msg: matrix B restoration in ralg solver
>   4750  6.292e+09               4.14
>   4800  2.558e+09               3.96
>   4850  2.898e+09               4.53
>   4900  1.581e+09               4.21
>   4950  1.272e+09               4.28
>   5000  5.860e+09               6.34
>   5050  4.163e+09               4.56
>   5100  3.478e+09               4.22
>   5150  3.238e+09               4.31
>   5200  2.862e+09               3.92
> OO info:  debug msg: matrix B restoration in ralg solver
>   5250  3.259e+09               4.36
>   5300  2.207e+09               3.91
>   5350  1.760e+09               3.74
>   5400  1.560e+09               3.93
>   5450  1.925e+09               4.41
>   5500  1.739e+09               4.41
>   5550  1.640e+09               4.42
>   5600  8.408e+10               4.93
>   5650  9.792e+10               4.69
>   5700  1.303e+11               4.75
>   5750  2.450e+11               5.44
>   5800  4.913e+11               5.33
>   5850  2.536e+11               6.00
>   5900  3.098e+11               5.70
>   5950  8.987e+10               5.37
> OO info:  debug msg: matrix B restoration in ralg solver
>   6000  1.037e+12               6.00
>   6050  3.448e+11               8.99
>   6100  8.307e+12               6.40
>   6150  1.589e+12               5.87
>   6200  1.213e+12               5.27
> OO info:  debug msg: matrix B restoration in ralg solver
>   6250  1.224e+12               5.45
>   6300  7.495e+11               5.00
>   6350  3.998e+11              15.67
>   6400  3.987e+11               5.00
>   6450  3.127e+11               5.02
>   6500  2.419e+11               5.27
>   6550  3.691e+11               5.13
>   6600  6.414e+11               5.74
>   6650  1.329e+12               5.92
>   6700  3.528e+11               5.18
>   6750  2.981e+11               4.78
>   6800  5.060e+11               5.51
>   6850  4.760e+11               5.09
>   6900  4.499e+11               5.10
>   6950  1.069e+12               5.86
>   7000  6.326e+11               5.26
>   7050  5.217e+11               5.18
>   7100  5.029e+11               5.16
>   7150  8.043e+12               6.43
>   7200  1.073e+13               6.51
>   7250  2.658e+12               6.18
>   7300  2.053e+11               4.81
>   7350  1.040e+12               5.45
>   7400  2.030e+12               6.08
>   7450  2.131e+12               6.11
>   7500  3.493e+11               5.17
>   7550  2.420e+11               5.04
>   7600  2.344e+12               6.17
>   7650  3.515e+11               5.62
>   7700  2.135e+11               5.35
>   7750  1.411e+11               4.78
>   7800  8.295e+12               6.46
>   7850  7.406e+12               6.39
>   7900  9.030e+12               6.45
>   7950  1.677e+12               6.04
> OO info:  debug msg: matrix B restoration in ralg solver
>   8000  3.579e+12               6.23
>   8050  1.109e+12              10.92
>   8100  5.111e+12               5.80
>   8150  7.521e+12               6.08
>   8200  7.199e+12               5.85
> OO info:  debug msg: matrix B restoration in ralg solver
>   8250  7.812e+12               6.05
>   8300  5.366e+12               8.57
>   8350  5.689e+12               5.97
>   8400  5.140e+12               5.97
>   8450  3.909e+12               5.38
> OO info:  debug msg: matrix B restoration in ralg solver
>   8500  5.130e+12               6.12
>   8550  3.753e+12               6.36
>   8600  2.963e+12               5.43
>   8650  2.528e+12               5.44
>   8700  2.134e+12               5.46
> OO info:  debug msg: matrix B restoration in ralg solver
>   8750  1.760e+12               5.46
>   8800  1.467e+12               5.27
>   8850  2.764e+12              12.53
>   8900  2.152e+12               5.63
>   8950  2.532e+12               5.86
> OO info:  debug msg: matrix B restoration in ralg solver
>   9000  1.884e+12               5.67
>   9050  4.073e+12              12.35
>   9100  1.709e+12               5.38
>   9150  1.398e+12               5.57
>   9200  1.248e+12               5.60
> OO info:  debug msg: matrix B restoration in ralg solver
>   9250  1.044e+12               5.14
>   9300  7.844e+11               5.21
>   9350  6.360e+11               5.47
>   9400  6.253e+11               5.67
>   9450  3.557e+11               4.91
>   9500  3.400e+11               5.29
>   9550  3.160e+11               5.30
>   9600  2.601e+11               4.94
>   9650  2.199e+11               4.85
>   9700  5.335e+12              13.48
> OO info:  debug msg: matrix B restoration in ralg solver
>   9750  5.933e+12               6.24
>   9800  4.174e+12               8.76
>   9850  3.803e+12               5.52
>   9900  2.854e+12               5.50
>   9950  2.014e+12               5.47
> 10000  3.285e+12               6.13
> 10001  3.285e+12               6.13
> istop:  -7 (Max Iter has been reached)
> Solver:   Time Elapsed = 56.05  CPU Time Elapsed = 31.82
> Plotting: Time Elapsed = 62.35  CPU Time Elapsed = 32.57
> NO FEASIBLE SOLUTION is obtained (max residual = 1.4e+06, 
> objFunc = 3.2852899e+12)
> _______________________________________________
> 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 Sep  5 15:53:05 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 05 Sep 2008 21:53:05 +0200
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C18B1C.4030801@scipy.org>
References: <48C14EEC.8060300@relativita.com> <48C155BF.7080008@relativita.com>
	<48C16C52.1040101@scipy.org> <web-110798387@uni-stuttgart.de>
	<48C18B1C.4030801@scipy.org>
Message-ID: <web-110801112@uni-stuttgart.de>

On Fri, 05 Sep 2008 22:40:12 +0300
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> Hi Nils,
> after some my modifications the file nlp_1.py become 
>hard to be solved 
> by any solver (I mean connected to OO) - you can try 
>solving it by 
> algencan or ipopt and see results (output).
> 
> So I have committed some changes to nlp_1.py. As for 
>other tests (like 
> nlp_bench_1, nlp_3) they work ok (nlp_2 for ralg 
>requires p.maxIter = 2000).
> 
> Regards, D.
> 
> Nils Wagner wrote:
>> On Fri, 05 Sep 2008 20:28:50 +0300
>>   dmitrey <dmitrey.kroshko at scipy.org> wrote:
>>   
>>> Hi Emanuele,
>>> as it is mentioned in openopt install webpage and 
>>> README.txt  numpy v 
>>>     
>>>> = 1.1.0 is recommended. Some other oo users informed of 
>>>>       
>>> bugs due to 
>>> older versions.
>>>
>>> Could you inform what will be outputed if you set 
>>> p.debug = 1? (either 
>>> directly or via p = NLP(..., debug=1,...))
>>>
>>> If the problem with numpy versions is critical for users 
>>> of your soft, 
>>> you'd better to put more recent numpy into Debian soft 
>>> channel.
>>>
>>> Regards, D.
>>>
>>> Emanuele Olivetti wrote:
>>>     
>>>> Same problem with numpy 1.0.4 + scipy 0.6.0
>>>> (shipped with ubuntu 8.04 hardy heron).
>>>>
>>>> E.
>>>>
>>>> Emanuele Olivetti wrote:
>>>>   
>>>>       
>>>>> Dear all and Dmitrey,
>>>>>
>>>>> I've just updated to latest openopt (SVN). When using 
>>>>> numpy 1.0.3
>>>>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) 
>>>>> openopt says
>>>>> that "ralg" (NLP) algorithm is missing! With more recent 
>>>>> numpy
>>>>> and scipy it seems to work reliably. But what happened 
>>>>> with respect
>>>>> to older numpy+scipy? In that case even running 
>>>>> examples/nlp_1.py
>>>>> returns:
>>>>> ----
>>>>> $ python nlp_1.py
>>>>> OpenOpt checks user-supplied gradient df (shape: (150,) 
>>>>> )
>>>>> according to:
>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>>> = 0.01
>>>>> derivatives are equal
>>>>> ========================
>>>>> OpenOpt checks user-supplied gradient dc (shape: (2, 
>>>>> 150) )
>>>>> according to:
>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>>> = 0.01
>>>>> derivatives are equal
>>>>> ========================
>>>>> OpenOpt checks user-supplied gradient dh (shape: (2, 
>>>>> 150) )
>>>>> according to:
>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation 
>>>>> = 0.01
>>>>> derivatives are equal
>>>>> ========================
>>>>> OO Error:incorrect solver is called, maybe the solver 
>>>>> "ralg" is not
>>>>> installed. Maybe setting p.debug=1 could specify the 
>>>>> matter more precisely
>>>>> Traceback (most recent call last):
>>>>>   File "nlp_1.py", line 110, in <module>
>>>>>     r = p.solve('ralg')
>>>>>   File
>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>>>> line 185, in solve
>>>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>>>   File
>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>>>> line 48, in runProbSolver
>>>>>     p.err('incorrect solver is called, maybe the solver 
>>>>> "' + solver_str
>>>>> +'" is not installed. Maybe setting p.debug=1 could 
>>>>> specify the matter
>>>>> more precisely')
>>>>>   File
>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>>>>> line 16, in ooerr
>>>>>     raise OpenOptException(msg)
>>>>> scikits.openopt.Kernel.oologfcn.OpenOptException: 
>>>>> incorrect solver is
>>>>> called, maybe the solver "ralg" is not installed. Maybe 
>>>>> setting
>>>>> p.debug=1 could specify the matter more precisely
>>>>> ----
>>>>>
>>>>> This did not happen before so I guess it is due to a 
>>>>> recent
>>>>> commit. It is possible to solve the problem?
>>>>>
>>>>> Kind Regards,
>>>>>
>>>>> Emanuele
>>>>>
>>>>> _______________________________________________
>>>>> SciPy-user mailing list
>>>>> SciPy-user at scipy.org
>>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>>
>>>>>   
>>>>>     
>>>>>         
>>>> _______________________________________________
>>>> SciPy-user mailing list
>>>> SciPy-user at scipy.org
>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>
>>>>
>>>>
>>>>   
>>>>       
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>     
>>
>>   
>> Dmitrey,
>>
>> I am using
>>   
>>>>> numpy.__version__
>>>>>         
>> '1.3.0.dev5790'
>>
>> Cheers,
>>           Nils
>>
>> Here comes the output of nlp_1.py:
>>
>> OpenOpt checks user-supplied gradient df (shape: (150,) 
>>)
>> according to:
>>      prob.diffInt = [  1.00000000e-07]
>>      |1 - info_user/info_numerical| <= prob.maxViolation 
>>= 
>> 0.01
>> derivatives are equal
>> ========================
>> OpenOpt checks user-supplied gradient dc (shape: (2, 
>>150) 
>> )
>> according to:
>>      prob.diffInt = [  1.00000000e-07]
>>      |1 - info_user/info_numerical| <= prob.maxViolation 
>>= 
>> 0.01
>> derivatives are equal
>> ========================
>> OpenOpt checks user-supplied gradient dh (shape: (2, 
>>150) 
>> )
>> according to:
>>      prob.diffInt = [  1.00000000e-07]
>>      |1 - info_user/info_numerical| <= prob.maxViolation 
>>= 
>> 0.01
>> derivatives are equal
>> ========================
>> -----------------------------------------------------
>> solver: ralg   problem: unnamed   goal: minimum
>>   iter    objFunVal    log10(maxResidual)
>>      0  8.596e+03               3.91
>> OpenOpt debug msg:  hs: 4.0
>> OpenOpt debug msg:  ls: 2
>>     50  2.800e+03               0.79
>>    100  1.754e+03               0.52
>>    150  9.075e+02               0.31
>>    200  4.455e+02              -0.03
>>    250  3.682e+02              -0.48
>>    300  3.465e+02              -1.15
>>    350  3.409e+02              -1.81
>>    400  1.911e+02              -3.14
>>    450  1.373e+02              -3.07
>> OO info:  debug msg: matrix B restoration in ralg solver
>>    500  1.065e+03               1.20
>>    550  2.224e+03               1.21
>>    600  1.822e+03               0.43
>>    650  2.178e+03               0.45
>>    700  2.576e+03               0.48
>>    750  2.840e+03               0.53
>>    800  3.068e+03               0.59
>>    850  7.958e+03               1.37
>>    900  2.174e+04               1.54
>>    950  3.341e+04               1.37
>>   1000  7.463e+04               2.17
>>   1050  3.692e+05               2.50
>>   1100  1.940e+05               2.16
>>   1150  1.482e+05               1.77
>>   1200  1.719e+05               1.86
>>   1250  2.963e+05               2.52
>>   1300  1.603e+05               2.27
>>   1350  2.299e+05               2.56
>>   1400  3.243e+05               2.63
>>   1450  2.663e+05               2.51
>>   1500  3.064e+05               2.55
>>   1550  4.297e+05               2.74
>>   1600  1.629e+05               2.80
>>   1650  2.379e+05               2.33
>>   1700  2.086e+05               2.28
>>   1750  1.214e+05               2.22
>>   1800  4.913e+04               1.58
>>   1850  3.862e+04               1.65
>>   1900  1.610e+05               2.53
>>   1950  3.576e+04               1.44
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   2000  7.286e+05               2.42
>>   2050  5.268e+05               2.50
>>   2100  1.403e+05               2.01
>>   2150  1.029e+05               1.96
>>   2200  9.997e+04               2.15
>>   2250  7.424e+05               2.92
>>   2300  5.514e+04               1.55
>>   2350  2.518e+05               2.66
>>   2400  5.051e+04               1.78
>>   2450  5.006e+04               2.05
>>   2500  4.279e+04               1.44
>>   2550  4.509e+04               1.62
>>   2600  1.331e+05               2.45
>>   2650  4.061e+04               1.41
>>   2700  5.198e+04               1.90
>>   2750  3.489e+09               4.77
>>   2800  6.938e+04               2.22
>>   2850  2.474e+10               5.20
>>   2900  4.259e+07               3.81
>>   2950  1.044e+05               2.40
>>   3000  6.411e+10               5.40
>>   3050  6.232e+07               3.89
>>   3100  1.830e+06               3.13
>>   3150  4.635e+04               1.45
>>   3200  1.770e+09               4.62
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   3250  1.764e+11               5.57
>>   3300  3.792e+09               4.01
>>   3350  1.554e+10               5.05
>>   3400  7.294e+09               4.81
>>   3450  7.227e+09               4.81
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   3500  1.415e+11               5.56
>>   3550  1.795e+10               6.16
>>   3600  5.205e+09               4.40
>>   3650  1.641e+10               5.04
>>   3700  1.408e+10               5.01
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   3750  1.277e+10               4.96
>>   3800  5.576e+09               3.89
>>   3850  5.008e+09               3.97
>>   3900  4.475e+09               4.04
>>   3950  3.993e+09               4.11
>>   4000  3.558e+09               4.17
>>   4050  3.237e+09               4.24
>>   4100  2.844e+09               4.24
>>   4150  1.077e+10               4.83
>>   4200  9.891e+09               4.82
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   4250  4.720e+09               4.12
>>   4300  3.411e+09               4.02
>>   4350  1.768e+09               6.43
>>   4400  1.851e+09               4.31
>>   4450  1.448e+09               3.99
>>   4500  1.248e+09               3.84
>>   4550  1.099e+09               3.80
>>   4600  6.053e+09               4.85
>>   4650  8.905e+08               3.86
>>   4700  1.446e+09               4.43
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   4750  6.292e+09               4.14
>>   4800  2.558e+09               3.96
>>   4850  2.898e+09               4.53
>>   4900  1.581e+09               4.21
>>   4950  1.272e+09               4.28
>>   5000  5.860e+09               6.34
>>   5050  4.163e+09               4.56
>>   5100  3.478e+09               4.22
>>   5150  3.238e+09               4.31
>>   5200  2.862e+09               3.92
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   5250  3.259e+09               4.36
>>   5300  2.207e+09               3.91
>>   5350  1.760e+09               3.74
>>   5400  1.560e+09               3.93
>>   5450  1.925e+09               4.41
>>   5500  1.739e+09               4.41
>>   5550  1.640e+09               4.42
>>   5600  8.408e+10               4.93
>>   5650  9.792e+10               4.69
>>   5700  1.303e+11               4.75
>>   5750  2.450e+11               5.44
>>   5800  4.913e+11               5.33
>>   5850  2.536e+11               6.00
>>   5900  3.098e+11               5.70
>>   5950  8.987e+10               5.37
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   6000  1.037e+12               6.00
>>   6050  3.448e+11               8.99
>>   6100  8.307e+12               6.40
>>   6150  1.589e+12               5.87
>>   6200  1.213e+12               5.27
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   6250  1.224e+12               5.45
>>   6300  7.495e+11               5.00
>>   6350  3.998e+11              15.67
>>   6400  3.987e+11               5.00
>>   6450  3.127e+11               5.02
>>   6500  2.419e+11               5.27
>>   6550  3.691e+11               5.13
>>   6600  6.414e+11               5.74
>>   6650  1.329e+12               5.92
>>   6700  3.528e+11               5.18
>>   6750  2.981e+11               4.78
>>   6800  5.060e+11               5.51
>>   6850  4.760e+11               5.09
>>   6900  4.499e+11               5.10
>>   6950  1.069e+12               5.86
>>   7000  6.326e+11               5.26
>>   7050  5.217e+11               5.18
>>   7100  5.029e+11               5.16
>>   7150  8.043e+12               6.43
>>   7200  1.073e+13               6.51
>>   7250  2.658e+12               6.18
>>   7300  2.053e+11               4.81
>>   7350  1.040e+12               5.45
>>   7400  2.030e+12               6.08
>>   7450  2.131e+12               6.11
>>   7500  3.493e+11               5.17
>>   7550  2.420e+11               5.04
>>   7600  2.344e+12               6.17
>>   7650  3.515e+11               5.62
>>   7700  2.135e+11               5.35
>>   7750  1.411e+11               4.78
>>   7800  8.295e+12               6.46
>>   7850  7.406e+12               6.39
>>   7900  9.030e+12               6.45
>>   7950  1.677e+12               6.04
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   8000  3.579e+12               6.23
>>   8050  1.109e+12              10.92
>>   8100  5.111e+12               5.80
>>   8150  7.521e+12               6.08
>>   8200  7.199e+12               5.85
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   8250  7.812e+12               6.05
>>   8300  5.366e+12               8.57
>>   8350  5.689e+12               5.97
>>   8400  5.140e+12               5.97
>>   8450  3.909e+12               5.38
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   8500  5.130e+12               6.12
>>   8550  3.753e+12               6.36
>>   8600  2.963e+12               5.43
>>   8650  2.528e+12               5.44
>>   8700  2.134e+12               5.46
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   8750  1.760e+12               5.46
>>   8800  1.467e+12               5.27
>>   8850  2.764e+12              12.53
>>   8900  2.152e+12               5.63
>>   8950  2.532e+12               5.86
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   9000  1.884e+12               5.67
>>   9050  4.073e+12              12.35
>>   9100  1.709e+12               5.38
>>   9150  1.398e+12               5.57
>>   9200  1.248e+12               5.60
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   9250  1.044e+12               5.14
>>   9300  7.844e+11               5.21
>>   9350  6.360e+11               5.47
>>   9400  6.253e+11               5.67
>>   9450  3.557e+11               4.91
>>   9500  3.400e+11               5.29
>>   9550  3.160e+11               5.30
>>   9600  2.601e+11               4.94
>>   9650  2.199e+11               4.85
>>   9700  5.335e+12              13.48
>> OO info:  debug msg: matrix B restoration in ralg solver
>>   9750  5.933e+12               6.24
>>   9800  4.174e+12               8.76
>>   9850  3.803e+12               5.52
>>   9900  2.854e+12               5.50
>>   9950  2.014e+12               5.47
>> 10000  3.285e+12               6.13
>> 10001  3.285e+12               6.13
>> istop:  -7 (Max Iter has been reached)
>> Solver:   Time Elapsed = 56.05  CPU Time Elapsed = 31.82
>> Plotting: Time Elapsed = 62.35  CPU Time Elapsed = 32.57
>> NO FEASIBLE SOLUTION is obtained (max residual = 
>>1.4e+06, 
>> objFunc = 3.2852899e+12)
>> _______________________________________________
>> 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
  
Now it works for me
------------------------------------------------------------------------
r1270 | dmitrey.kroshko | 2008-09-05 21:33:49 +0200 (Fri, 
05 Sep 2008) | 1 line

some changes in nlp_1.py


OpenOpt checks user-supplied gradient df (shape: (150,) )
according to:
     prob.diffInt = [  1.00000000e-07]
     |1 - info_user/info_numerical| <= prob.maxViolation = 
0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dc (shape: (2, 150) 
)
according to:
     prob.diffInt = [  1.00000000e-07]
     |1 - info_user/info_numerical| <= prob.maxViolation = 
0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dh (shape: (2, 150) 
)
according to:
     prob.diffInt = [  1.00000000e-07]
     |1 - info_user/info_numerical| <= prob.maxViolation = 
0.01
derivatives are equal
========================
-----------------------------------------------------
solver: ralg   problem: unnamed   goal: minimum
  iter    objFunVal    log10(maxResidual)
     0  8.596e+03               5.73
OpenOpt debug msg:  hs: 16.0
OpenOpt debug msg:  ls: 4
    50  5.237e+03               1.08
   100  7.347e+03               1.04
   150  2.248e+04               1.24
   200  7.588e+03               1.24
   250  3.281e+03               0.74
   300  2.780e+03               0.59
   350  2.328e+03               0.52
   400  1.748e+03               0.39
   450  1.433e+03               0.27
   500  9.347e+02               0.10
   550  5.696e+02              -0.17
   600  4.870e+02              -0.46
   650  3.879e+02              -0.84
   700  3.319e+02              -1.35
   750  1.433e+02              -1.42
   800  1.444e+02              -1.46
   850  1.380e+02              -3.10
   900  1.337e+02              -3.03
   950  1.294e+02              -3.10
OO info:  debug msg: matrix B restoration in ralg solver
  1000  1.282e+02              -3.10
  1050  1.281e+02              -3.10
  1100  1.281e+02              -2.91
  1135  1.281e+02              -3.10
/usr/local/lib64/python2.5/site-packages/matplotlib/axes.py:4827: 
DeprecationWarning: replace "faceted=False" with 
"edgecolors='none'"
   DeprecationWarning)   #2008/04/18
istop:  3 (|| X[k] - X[k-1] || < xtol)
Solver:   Time Elapsed = 7.71   CPU Time Elapsed = 5.33
Plotting: Time Elapsed = 13.29  CPU Time Elapsed = 7.63
objFunValue: 128.08949 (feasible, max constraint = 
 0.0008)

Cheers,
             Nils


From emanuele at relativita.com  Fri Sep  5 18:51:01 2008
From: emanuele at relativita.com (Emanuele Olivetti)
Date: Sat, 06 Sep 2008 00:51:01 +0200
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C16C52.1040101@scipy.org>
References: <48C14EEC.8060300@relativita.com>
	<48C155BF.7080008@relativita.com> <48C16C52.1040101@scipy.org>
Message-ID: <48C1B7D5.9060006@relativita.com>

Thanks for help.

Unfortunately after updating from SVN again I'm not able to reproduce
the same error, but instead I get this, about failing to import "cond" from
numpy.linalg. It seems that "cond" is not available until the very latest
numpy (and even autogenerated NumPy API on scipy.org have no "cond"):
----
$ python openopt/scikits/openopt/examples/nlp_1.py
OpenOpt checks user-supplied gradient df (shape: (150,) )
according to:
    prob.diffInt = [  1.00000000e-07]
    |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
according to:
    prob.diffInt = [  1.00000000e-07]
    |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
derivatives are equal
========================
OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
according to:
    prob.diffInt = [  1.00000000e-07]
    |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
derivatives are equal
========================
Traceback (most recent call last):
  File "openopt/scikits/openopt/examples/nlp_1.py", line 108, in <module>
    r = p.solve('ralg', debug = 1)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
line 185, in solve
    return runProbSolver(self, solvers, *args, **kwargs)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
line 43, in runProbSolver
    solverClass =  getattr(my_import(__solverPaths__[solver_str]),
solver_str)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
line 268, in my_import
    mod = __import__(name)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/solvers/UkrOpt/ralg_oo.py",
line 2, in <module>
    from numpy.linalg import norm, cond
ImportError: cannot import name cond
----

Any suggestion on how to solve this?

Sorry for the mess.

Emanuele

Note: this error pops out using numpy+scipy shipped with
ubuntu. When using recent SVN version of numpy+scipy
everything works well.

dmitrey wrote:
> Hi Emanuele,
>  as it is mentioned in openopt install webpage and README.txt  numpy v 
>  >= 1.1.0 is recommended. Some other oo users informed of bugs due to 
> older versions.
>
> Could you inform what will be outputed if you set p.debug = 1? (either 
> directly or via p = NLP(..., debug=1,...))
>
> If the problem with numpy versions is critical for users of your soft, 
> you'd better to put more recent numpy into Debian soft channel.
>
> Regards, D.
>
> Emanuele Olivetti wrote:
>   
>> Same problem with numpy 1.0.4 + scipy 0.6.0
>> (shipped with ubuntu 8.04 hardy heron).
>>
>> E.
>>
>> Emanuele Olivetti wrote:
>>   
>>     
>>> Dear all and Dmitrey,
>>>
>>> I've just updated to latest openopt (SVN). When using numpy 1.0.3
>>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) openopt says
>>> that "ralg" (NLP) algorithm is missing! With more recent numpy
>>> and scipy it seems to work reliably. But what happened with respect
>>> to older numpy+scipy? In that case even running examples/nlp_1.py
>>> returns:
>>> ----
>>> $ python nlp_1.py
>>> OpenOpt checks user-supplied gradient df (shape: (150,) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>> derivatives are equal
>>> ========================
>>> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>> derivatives are equal
>>> ========================
>>> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>> derivatives are equal
>>> ========================
>>> OO Error:incorrect solver is called, maybe the solver "ralg" is not
>>> installed. Maybe setting p.debug=1 could specify the matter more precisely
>>> Traceback (most recent call last):
>>>   File "nlp_1.py", line 110, in <module>
>>>     r = p.solve('ralg')
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>> line 185, in solve
>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>> line 48, in runProbSolver
>>>     p.err('incorrect solver is called, maybe the solver "' + solver_str
>>> +'" is not installed. Maybe setting p.debug=1 could specify the matter
>>> more precisely')
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>>> line 16, in ooerr
>>>     raise OpenOptException(msg)
>>> scikits.openopt.Kernel.oologfcn.OpenOptException: incorrect solver is
>>> called, maybe the solver "ralg" is not installed. Maybe setting
>>> p.debug=1 could specify the matter more precisely
>>> ----
>>>
>>> This did not happen before so I guess it is due to a recent
>>> commit. It is possible to solve the problem?
>>>
>>> Kind Regards,
>>>
>>> Emanuele
>>>
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>
>>>   
>>>     
>>>       
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>>
>>
>>   
>>     
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   



From emanuele at relativita.com  Fri Sep  5 19:00:17 2008
From: emanuele at relativita.com (Emanuele Olivetti)
Date: Sat, 06 Sep 2008 01:00:17 +0200
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C1B7D5.9060006@relativita.com>
References: <48C14EEC.8060300@relativita.com>	<48C155BF.7080008@relativita.com>
	<48C16C52.1040101@scipy.org> <48C1B7D5.9060006@relativita.com>
Message-ID: <48C1BA01.4000501@relativita.com>

OK. Running another custom example I got again the initial "ralg missing"
error message. Increasing verbosity as you suggested (problem.debug = 1)
shows the same error message shown before, i.e. "cond" is not available
in numpy.linalg, so import fails:
----
...<cut>...
 in solve
    result = self.problem.solve(self.optimization_algorithm) # perform
optimization!
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
line 185, in solve
    return runProbSolver(self, solvers, *args, **kwargs)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
line 43, in runProbSolver
    solverClass =  getattr(my_import(__solverPaths__[solver_str]),
solver_str)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
line 268, in my_import
    mod = __import__(name)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/solvers/UkrOpt/ralg_oo.py",
line 2, in <module>
    from numpy.linalg import norm, cond
ImportError: cannot import name cond
----

Hope this helps,

Emanuele

Emanuele Olivetti wrote:
> Thanks for help.
>
> Unfortunately after updating from SVN again I'm not able to reproduce
> the same error, but instead I get this, about failing to import "cond" from
> numpy.linalg. It seems that "cond" is not available until the very latest
> numpy (and even autogenerated NumPy API on scipy.org have no "cond"):
> ----
> $ python openopt/scikits/openopt/examples/nlp_1.py
> OpenOpt checks user-supplied gradient df (shape: (150,) )
> according to:
>     prob.diffInt = [  1.00000000e-07]
>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
> derivatives are equal
> ========================
> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
> according to:
>     prob.diffInt = [  1.00000000e-07]
>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
> derivatives are equal
> ========================
> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
> according to:
>     prob.diffInt = [  1.00000000e-07]
>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
> derivatives are equal
> ========================
> Traceback (most recent call last):
>   File "openopt/scikits/openopt/examples/nlp_1.py", line 108, in <module>
>     r = p.solve('ralg', debug = 1)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
> line 185, in solve
>     return runProbSolver(self, solvers, *args, **kwargs)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
> line 43, in runProbSolver
>     solverClass =  getattr(my_import(__solverPaths__[solver_str]),
> solver_str)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
> line 268, in my_import
>     mod = __import__(name)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/solvers/UkrOpt/ralg_oo.py",
> line 2, in <module>
>     from numpy.linalg import norm, cond
> ImportError: cannot import name cond
> ----
>
> Any suggestion on how to solve this?
>
> Sorry for the mess.
>
> Emanuele
>
> Note: this error pops out using numpy+scipy shipped with
> ubuntu. When using recent SVN version of numpy+scipy
> everything works well.
>
> dmitrey wrote:
>   
>> Hi Emanuele,
>>  as it is mentioned in openopt install webpage and README.txt  numpy v 
>>  >= 1.1.0 is recommended. Some other oo users informed of bugs due to 
>> older versions.
>>
>> Could you inform what will be outputed if you set p.debug = 1? (either 
>> directly or via p = NLP(..., debug=1,...))
>>
>> If the problem with numpy versions is critical for users of your soft, 
>> you'd better to put more recent numpy into Debian soft channel.
>>
>> Regards, D.
>>
>> Emanuele Olivetti wrote:
>>   
>>     
>>> Same problem with numpy 1.0.4 + scipy 0.6.0
>>> (shipped with ubuntu 8.04 hardy heron).
>>>
>>> E.
>>>
>>> Emanuele Olivetti wrote:
>>>   
>>>     
>>>       
>>>> Dear all and Dmitrey,
>>>>
>>>> I've just updated to latest openopt (SVN). When using numpy 1.0.3
>>>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) openopt says
>>>> that "ralg" (NLP) algorithm is missing! With more recent numpy
>>>> and scipy it seems to work reliably. But what happened with respect
>>>> to older numpy+scipy? In that case even running examples/nlp_1.py
>>>> returns:
>>>> ----
>>>> $ python nlp_1.py
>>>> OpenOpt checks user-supplied gradient df (shape: (150,) )
>>>> according to:
>>>>     prob.diffInt = [  1.00000000e-07]
>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>> derivatives are equal
>>>> ========================
>>>> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
>>>> according to:
>>>>     prob.diffInt = [  1.00000000e-07]
>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>> derivatives are equal
>>>> ========================
>>>> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
>>>> according to:
>>>>     prob.diffInt = [  1.00000000e-07]
>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>> derivatives are equal
>>>> ========================
>>>> OO Error:incorrect solver is called, maybe the solver "ralg" is not
>>>> installed. Maybe setting p.debug=1 could specify the matter more precisely
>>>> Traceback (most recent call last):
>>>>   File "nlp_1.py", line 110, in <module>
>>>>     r = p.solve('ralg')
>>>>   File
>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>>> line 185, in solve
>>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>>   File
>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>>> line 48, in runProbSolver
>>>>     p.err('incorrect solver is called, maybe the solver "' + solver_str
>>>> +'" is not installed. Maybe setting p.debug=1 could specify the matter
>>>> more precisely')
>>>>   File
>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>>>> line 16, in ooerr
>>>>     raise OpenOptException(msg)
>>>> scikits.openopt.Kernel.oologfcn.OpenOptException: incorrect solver is
>>>> called, maybe the solver "ralg" is not installed. Maybe setting
>>>> p.debug=1 could specify the matter more precisely
>>>> ----
>>>>
>>>> This did not happen before so I guess it is due to a recent
>>>> commit. It is possible to solve the problem?
>>>>
>>>> Kind Regards,
>>>>
>>>> Emanuele
>>>>
>>>> _______________________________________________
>>>> SciPy-user mailing list
>>>> SciPy-user at scipy.org
>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>
>>>>   
>>>>     
>>>>       
>>>>         
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>
>>>
>>>
>>>   
>>>     
>>>       
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>>   
>>     
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   



From dmitrey.kroshko at scipy.org  Sat Sep  6 03:16:41 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Sat, 06 Sep 2008 10:16:41 +0300
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C1BA01.4000501@relativita.com>
References: <48C14EEC.8060300@relativita.com>	<48C155BF.7080008@relativita.com>	<48C16C52.1040101@scipy.org>
	<48C1B7D5.9060006@relativita.com> <48C1BA01.4000501@relativita.com>
Message-ID: <48C22E59.8030208@scipy.org>

Hi Emanuele,
update svn and try now,
HTH, D.

P.S. IIRC you are deal with box-bounded problems, let me remember you 
once again, that ralg (especially current implementation) handles it 
very badly (especially when lots of active constraints in optim point) 
in comparison to scipy_lbfgsb, scipy_tnc or algencan (all are available 
from oo, requires scipy or algencan installed), these ones have very 
appropriate specialized box-bound solvers.
Regards, D

Emanuele Olivetti wrote:
> OK. Running another custom example I got again the initial "ralg missing"
> error message. Increasing verbosity as you suggested (problem.debug = 1)
> shows the same error message shown before, i.e. "cond" is not available
> in numpy.linalg, so import fails:
> ----
> ...<cut>...
>  in solve
>     result = self.problem.solve(self.optimization_algorithm) # perform
> optimization!
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
> line 185, in solve
>     return runProbSolver(self, solvers, *args, **kwargs)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
> line 43, in runProbSolver
>     solverClass =  getattr(my_import(__solverPaths__[solver_str]),
> solver_str)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
> line 268, in my_import
>     mod = __import__(name)
>   File
> "/usr/lib/python2.5/site-packages/scikits/openopt/solvers/UkrOpt/ralg_oo.py",
> line 2, in <module>
>     from numpy.linalg import norm, cond
> ImportError: cannot import name cond
> ----
>
> Hope this helps,
>
> Emanuele
>
> Emanuele Olivetti wrote:
>   
>> Thanks for help.
>>
>> Unfortunately after updating from SVN again I'm not able to reproduce
>> the same error, but instead I get this, about failing to import "cond" from
>> numpy.linalg. It seems that "cond" is not available until the very latest
>> numpy (and even autogenerated NumPy API on scipy.org have no "cond"):
>> ----
>> $ python openopt/scikits/openopt/examples/nlp_1.py
>> OpenOpt checks user-supplied gradient df (shape: (150,) )
>> according to:
>>     prob.diffInt = [  1.00000000e-07]
>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>> derivatives are equal
>> ========================
>> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
>> according to:
>>     prob.diffInt = [  1.00000000e-07]
>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>> derivatives are equal
>> ========================
>> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
>> according to:
>>     prob.diffInt = [  1.00000000e-07]
>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>> derivatives are equal
>> ========================
>> Traceback (most recent call last):
>>   File "openopt/scikits/openopt/examples/nlp_1.py", line 108, in <module>
>>     r = p.solve('ralg', debug = 1)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>> line 185, in solve
>>     return runProbSolver(self, solvers, *args, **kwargs)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>> line 43, in runProbSolver
>>     solverClass =  getattr(my_import(__solverPaths__[solver_str]),
>> solver_str)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>> line 268, in my_import
>>     mod = __import__(name)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/solvers/UkrOpt/ralg_oo.py",
>> line 2, in <module>
>>     from numpy.linalg import norm, cond
>> ImportError: cannot import name cond
>> ----
>>
>> Any suggestion on how to solve this?
>>
>> Sorry for the mess.
>>
>> Emanuele
>>
>> Note: this error pops out using numpy+scipy shipped with
>> ubuntu. When using recent SVN version of numpy+scipy
>> everything works well.
>>
>> dmitrey wrote:
>>   
>>     
>>> Hi Emanuele,
>>>  as it is mentioned in openopt install webpage and README.txt  numpy v 
>>>  >= 1.1.0 is recommended. Some other oo users informed of bugs due to 
>>> older versions.
>>>
>>> Could you inform what will be outputed if you set p.debug = 1? (either 
>>> directly or via p = NLP(..., debug=1,...))
>>>
>>> If the problem with numpy versions is critical for users of your soft, 
>>> you'd better to put more recent numpy into Debian soft channel.
>>>
>>> Regards, D.
>>>
>>> Emanuele Olivetti wrote:
>>>   
>>>     
>>>       
>>>> Same problem with numpy 1.0.4 + scipy 0.6.0
>>>> (shipped with ubuntu 8.04 hardy heron).
>>>>
>>>> E.
>>>>
>>>> Emanuele Olivetti wrote:
>>>>   
>>>>     
>>>>       
>>>>         
>>>>> Dear all and Dmitrey,
>>>>>
>>>>> I've just updated to latest openopt (SVN). When using numpy 1.0.3
>>>>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) openopt says
>>>>> that "ralg" (NLP) algorithm is missing! With more recent numpy
>>>>> and scipy it seems to work reliably. But what happened with respect
>>>>> to older numpy+scipy? In that case even running examples/nlp_1.py
>>>>> returns:
>>>>> ----
>>>>> $ python nlp_1.py
>>>>> OpenOpt checks user-supplied gradient df (shape: (150,) )
>>>>> according to:
>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>>> derivatives are equal
>>>>> ========================
>>>>> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
>>>>> according to:
>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>>> derivatives are equal
>>>>> ========================
>>>>> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
>>>>> according to:
>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>>> derivatives are equal
>>>>> ========================
>>>>> OO Error:incorrect solver is called, maybe the solver "ralg" is not
>>>>> installed. Maybe setting p.debug=1 could specify the matter more precisely
>>>>> Traceback (most recent call last):
>>>>>   File "nlp_1.py", line 110, in <module>
>>>>>     r = p.solve('ralg')
>>>>>   File
>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>>>> line 185, in solve
>>>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>>>   File
>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>>>> line 48, in runProbSolver
>>>>>     p.err('incorrect solver is called, maybe the solver "' + solver_str
>>>>> +'" is not installed. Maybe setting p.debug=1 could specify the matter
>>>>> more precisely')
>>>>>   File
>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>>>>> line 16, in ooerr
>>>>>     raise OpenOptException(msg)
>>>>> scikits.openopt.Kernel.oologfcn.OpenOptException: incorrect solver is
>>>>> called, maybe the solver "ralg" is not installed. Maybe setting
>>>>> p.debug=1 could specify the matter more precisely
>>>>> ----
>>>>>
>>>>> This did not happen before so I guess it is due to a recent
>>>>> commit. It is possible to solve the problem?
>>>>>
>>>>> Kind Regards,
>>>>>
>>>>> Emanuele
>>>>>
>>>>> _______________________________________________
>>>>> SciPy-user mailing list
>>>>> SciPy-user at scipy.org
>>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>>
>>>>>   
>>>>>     
>>>>>       
>>>>>         
>>>>>           
>>>> _______________________________________________
>>>> SciPy-user mailing list
>>>> SciPy-user at scipy.org
>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>
>>>>
>>>>
>>>>   
>>>>     
>>>>       
>>>>         
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>
>>>   
>>>     
>>>       
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>>   
>>     
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From emanuele at relativita.com  Sun Sep  7 03:47:40 2008
From: emanuele at relativita.com (Emanuele Olivetti)
Date: Sun, 07 Sep 2008 09:47:40 +0200
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C22E59.8030208@scipy.org>
References: <48C14EEC.8060300@relativita.com>	<48C155BF.7080008@relativita.com>	<48C16C52.1040101@scipy.org>	<48C1B7D5.9060006@relativita.com>
	<48C1BA01.4000501@relativita.com> <48C22E59.8030208@scipy.org>
Message-ID: <48C3871C.3000109@relativita.com>

Thanks Dmitrey,

I've just tried to update, install and test latest OpenOpt (SVN)
and it seems to work well with numpy 1.0.3 + scipy 0.5.2 shipped
with ubuntu 7.10 (or better, now nlp_1.py runs some
iterations and then the Python process gives "segmentation fault",
but this is a known Python/NumPy issue as far as I remember).
Anyway doesn't stop anymore. Thanks for the fix.

About your suggstion to use other solvers than ralg, I've
some good preliminary experience with scipy_lbfgsb; even
scipy_tnc seems to work well but requires a too recent scipy.
This is what happens with scipy 0.5.2 bundled with ubuntu 7.10
when calling scipy_tnc:
---
result = self.problem.solve(self.optimization_algorithm) # perform
optimization!
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
line 185, in solve
    return runProbSolver(self, solvers, *args, **kwargs)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
line 167, in runProbSolver
    solver(p)
  File
"/usr/lib/python2.5/site-packages/scikits/openopt/solvers/scipy_optim/scipy_tnc_oo.py",
line 38, in __solver__
    xf, nfeval, rc = fmin_tnc(p.f, x0 = p.x0, fprime=p.df, args=(),
approx_grad=0, bounds=bounds, messages=messages, maxfun=maxfun,
ftol=p.ftol, xtol=p.xtol, pgtol=p.gradtol)
TypeError: fmin_tnc() got an unexpected keyword argument 'xtol'
----
About algencan, I hanven't tried yet and I'm really interested. It is
just my lack of time and the external dependencies that slow down
my attempt. Anyway I've added optional logscale to my code, which
is similar to setting the lower bound I need:
problem.lb = N.zeros(problem.n) + contol

Best,

Emanuele


dmitrey wrote:
> Hi Emanuele,
> update svn and try now,
> HTH, D.
>
> P.S. IIRC you are deal with box-bounded problems, let me remember you 
> once again, that ralg (especially current implementation) handles it 
> very badly (especially when lots of active constraints in optim point) 
> in comparison to scipy_lbfgsb, scipy_tnc or algencan (all are available 
> from oo, requires scipy or algencan installed), these ones have very 
> appropriate specialized box-bound solvers.
> Regards, D
>
> Emanuele Olivetti wrote:
>   
>> OK. Running another custom example I got again the initial "ralg missing"
>> error message. Increasing verbosity as you suggested (problem.debug = 1)
>> shows the same error message shown before, i.e. "cond" is not available
>> in numpy.linalg, so import fails:
>> ----
>> ...<cut>...
>>  in solve
>>     result = self.problem.solve(self.optimization_algorithm) # perform
>> optimization!
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>> line 185, in solve
>>     return runProbSolver(self, solvers, *args, **kwargs)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>> line 43, in runProbSolver
>>     solverClass =  getattr(my_import(__solverPaths__[solver_str]),
>> solver_str)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>> line 268, in my_import
>>     mod = __import__(name)
>>   File
>> "/usr/lib/python2.5/site-packages/scikits/openopt/solvers/UkrOpt/ralg_oo.py",
>> line 2, in <module>
>>     from numpy.linalg import norm, cond
>> ImportError: cannot import name cond
>> ----
>>
>> Hope this helps,
>>
>> Emanuele
>>
>> Emanuele Olivetti wrote:
>>   
>>     
>>> Thanks for help.
>>>
>>> Unfortunately after updating from SVN again I'm not able to reproduce
>>> the same error, but instead I get this, about failing to import "cond" from
>>> numpy.linalg. It seems that "cond" is not available until the very latest
>>> numpy (and even autogenerated NumPy API on scipy.org have no "cond"):
>>> ----
>>> $ python openopt/scikits/openopt/examples/nlp_1.py
>>> OpenOpt checks user-supplied gradient df (shape: (150,) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>> derivatives are equal
>>> ========================
>>> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>> derivatives are equal
>>> ========================
>>> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
>>> according to:
>>>     prob.diffInt = [  1.00000000e-07]
>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>> derivatives are equal
>>> ========================
>>> Traceback (most recent call last):
>>>   File "openopt/scikits/openopt/examples/nlp_1.py", line 108, in <module>
>>>     r = p.solve('ralg', debug = 1)
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>> line 185, in solve
>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>> line 43, in runProbSolver
>>>     solverClass =  getattr(my_import(__solverPaths__[solver_str]),
>>> solver_str)
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>> line 268, in my_import
>>>     mod = __import__(name)
>>>   File
>>> "/usr/lib/python2.5/site-packages/scikits/openopt/solvers/UkrOpt/ralg_oo.py",
>>> line 2, in <module>
>>>     from numpy.linalg import norm, cond
>>> ImportError: cannot import name cond
>>> ----
>>>
>>> Any suggestion on how to solve this?
>>>
>>> Sorry for the mess.
>>>
>>> Emanuele
>>>
>>> Note: this error pops out using numpy+scipy shipped with
>>> ubuntu. When using recent SVN version of numpy+scipy
>>> everything works well.
>>>
>>> dmitrey wrote:
>>>   
>>>     
>>>       
>>>> Hi Emanuele,
>>>>  as it is mentioned in openopt install webpage and README.txt  numpy v 
>>>>  >= 1.1.0 is recommended. Some other oo users informed of bugs due to 
>>>> older versions.
>>>>
>>>> Could you inform what will be outputed if you set p.debug = 1? (either 
>>>> directly or via p = NLP(..., debug=1,...))
>>>>
>>>> If the problem with numpy versions is critical for users of your soft, 
>>>> you'd better to put more recent numpy into Debian soft channel.
>>>>
>>>> Regards, D.
>>>>
>>>> Emanuele Olivetti wrote:
>>>>   
>>>>     
>>>>       
>>>>         
>>>>> Same problem with numpy 1.0.4 + scipy 0.6.0
>>>>> (shipped with ubuntu 8.04 hardy heron).
>>>>>
>>>>> E.
>>>>>
>>>>> Emanuele Olivetti wrote:
>>>>>   
>>>>>     
>>>>>       
>>>>>         
>>>>>           
>>>>>> Dear all and Dmitrey,
>>>>>>
>>>>>> I've just updated to latest openopt (SVN). When using numpy 1.0.3
>>>>>> and scipy 0.5.2 (standard in Ubuntu 7.10 gutsy gibbon) openopt says
>>>>>> that "ralg" (NLP) algorithm is missing! With more recent numpy
>>>>>> and scipy it seems to work reliably. But what happened with respect
>>>>>> to older numpy+scipy? In that case even running examples/nlp_1.py
>>>>>> returns:
>>>>>> ----
>>>>>> $ python nlp_1.py
>>>>>> OpenOpt checks user-supplied gradient df (shape: (150,) )
>>>>>> according to:
>>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>>>> derivatives are equal
>>>>>> ========================
>>>>>> OpenOpt checks user-supplied gradient dc (shape: (2, 150) )
>>>>>> according to:
>>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>>>> derivatives are equal
>>>>>> ========================
>>>>>> OpenOpt checks user-supplied gradient dh (shape: (2, 150) )
>>>>>> according to:
>>>>>>     prob.diffInt = [  1.00000000e-07]
>>>>>>     |1 - info_user/info_numerical| <= prob.maxViolation = 0.01
>>>>>> derivatives are equal
>>>>>> ========================
>>>>>> OO Error:incorrect solver is called, maybe the solver "ralg" is not
>>>>>> installed. Maybe setting p.debug=1 could specify the matter more precisely
>>>>>> Traceback (most recent call last):
>>>>>>   File "nlp_1.py", line 110, in <module>
>>>>>>     r = p.solve('ralg')
>>>>>>   File
>>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/BaseProblem.py",
>>>>>> line 185, in solve
>>>>>>     return runProbSolver(self, solvers, *args, **kwargs)
>>>>>>   File
>>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/runProbSolver.py",
>>>>>> line 48, in runProbSolver
>>>>>>     p.err('incorrect solver is called, maybe the solver "' + solver_str
>>>>>> +'" is not installed. Maybe setting p.debug=1 could specify the matter
>>>>>> more precisely')
>>>>>>   File
>>>>>> "/usr/lib/python2.5/site-packages/scikits/openopt/Kernel/oologfcn.py",
>>>>>> line 16, in ooerr
>>>>>>     raise OpenOptException(msg)
>>>>>> scikits.openopt.Kernel.oologfcn.OpenOptException: incorrect solver is
>>>>>> called, maybe the solver "ralg" is not installed. Maybe setting
>>>>>> p.debug=1 could specify the matter more precisely
>>>>>> ----
>>>>>>
>>>>>> This did not happen before so I guess it is due to a recent
>>>>>> commit. It is possible to solve the problem?
>>>>>>
>>>>>> Kind Regards,
>>>>>>
>>>>>> Emanuele
>>>>>>
>>>>>> _______________________________________________
>>>>>> SciPy-user mailing list
>>>>>> SciPy-user at scipy.org
>>>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>>>
>>>>>>   
>>>>>>     
>>>>>>       
>>>>>>         
>>>>>>           
>>>>>>             
>>>>> _______________________________________________
>>>>> SciPy-user mailing list
>>>>> SciPy-user at scipy.org
>>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>>
>>>>>
>>>>>
>>>>>   
>>>>>     
>>>>>       
>>>>>         
>>>>>           
>>>> _______________________________________________
>>>> SciPy-user mailing list
>>>> SciPy-user at scipy.org
>>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>>
>>>>   
>>>>     
>>>>       
>>>>         
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>
>>>   
>>>     
>>>       
>> _______________________________________________
>> 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 dmitrey.kroshko at scipy.org  Sun Sep  7 04:10:04 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Sun, 07 Sep 2008 11:10:04 +0300
Subject: [SciPy-user] [OpenOpt] problem with ralg (latest SVN)
In-Reply-To: <48C3871C.3000109@relativita.com>
References: <48C14EEC.8060300@relativita.com>	<48C155BF.7080008@relativita.com>	<48C16C52.1040101@scipy.org>	<48C1B7D5.9060006@relativita.com>	<48C1BA01.4000501@relativita.com>
	<48C22E59.8030208@scipy.org> <48C3871C.3000109@relativita.com>
Message-ID: <48C38C5C.1080202@scipy.org>

Emanuele Olivetti wrote:
> Thanks Dmitrey,
>
> I've just tried to update, install and test latest OpenOpt (SVN)
> and it seems to work well with numpy 1.0.3 + scipy 0.5.2 shipped
> with ubuntu 7.10 (or better, now nlp_1.py runs some
> iterations and then the Python process gives "segmentation fault",
> but this is a known Python/NumPy issue as far as I remember).
> Anyway doesn't stop anymore. Thanks for the fix.
>
> About your suggstion to use other solvers than ralg, I've
> some good preliminary experience with scipy_lbfgsb; even
> scipy_tnc seems to work well but requires a too recent scipy.
> This is what happens with scipy 0.5.2 bundled with ubuntu 7.10
> when calling scipy_tnc:
> ---
> TypeError: fmin_tnc() got an unexpected keyword argument 'xtol'
> ----
> About algencan, I hanven't tried yet and I'm really interested. It is
> just my lack of time and the external dependencies that slow down
> my attempt. Anyway I've added optional logscale to my code, which
> is similar to setting the lower bound I need:
> problem.lb = N.zeros(problem.n) + contol
>
> Best,
>
> Emanuele
AFAIK currently (latest subversion) there are no OO solvers where the 
adding contol to box-bounds matters. As for next OO release I intend to 
have it done at *September 15.*
Regards, D.


From gyromagnetic at gmail.com  Sun Sep  7 11:37:28 2008
From: gyromagnetic at gmail.com (gfunch)
Date: Sun, 7 Sep 2008 09:37:28 -0600
Subject: [SciPy-user] segmentation fault in scipy.test()
Message-ID: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>

Hi,
I have tried to build SciPy (svn) on an x86_64 system running Linux
CentOS5, but ran into a segmentation fault while running the tests.

I first compiled and installed Lapack and ATLAS per some instructions
I found the web. I then built and installed numpy (which tested fine)
and then built scipy.

What commands should I run, and output should I generate, to best
diagnose the problem, and perhaps seek your kind help?

Thanks.

-gyro


From david at ar.media.kyoto-u.ac.jp  Sun Sep  7 11:27:37 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 08 Sep 2008 00:27:37 +0900
Subject: [SciPy-user] segmentation fault in scipy.test()
In-Reply-To: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>
References: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>
Message-ID: <48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>

gfunch wrote:
> Hi,
> I have tried to build SciPy (svn) on an x86_64 system running Linux
> CentOS5, but ran into a segmentation fault while running the tests.
>
> I first compiled and installed Lapack and ATLAS per some instructions
> I found the web. I then built and installed numpy (which tested fine)
> and then built scipy.
>
> What commands should I run, and output should I generate, to best
> diagnose the problem, and perhaps seek your kind help?
>   

Hi Gyro,

    Sorry for the bug in scipy. Could you give us the output when you
run the tests (e.g. which test failed and segfaulted ?). Something
useful would be the build log (output when run python setup.py
build/install; we need the output when build from scratch, e.g. after
having removed the build directory).
    If you are familiar with gdb, something useful may be a backtrace,
but first, I would like to check whether it is not a problem related to
ATLAS/Lapack,

cheers,

David


From gyromagnetic at gmail.com  Sun Sep  7 12:15:22 2008
From: gyromagnetic at gmail.com (gfunch)
Date: Sun, 7 Sep 2008 10:15:22 -0600
Subject: [SciPy-user] segmentation fault in scipy.test()
In-Reply-To: <48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>
References: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>
	<48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>
Message-ID: <ace26be40809070915k52b11fdfuef092c6a2483a333@mail.gmail.com>

On Sun, Sep 7, 2008 at 9:27 AM, David Cournapeau
<david at ar.media.kyoto-u.ac.jp> wrote:
> gfunch wrote:
>> Hi,
>> I have tried to build SciPy (svn) on an x86_64 system running Linux
>> CentOS5, but ran into a segmentation fault while running the tests.
>>
>> I first compiled and installed Lapack and ATLAS per some instructions
>> I found the web. I then built and installed numpy (which tested fine)
>> and then built scipy.
>>
>> What commands should I run, and output should I generate, to best
>> diagnose the problem, and perhaps seek your kind help?
>>
>
> Hi Gyro,
>
>    Sorry for the bug in scipy. Could you give us the output when you
> run the tests (e.g. which test failed and segfaulted ?). Something
> useful would be the build log (output when run python setup.py
> build/install; we need the output when build from scratch, e.g. after
> having removed the build directory).
>    If you are familiar with gdb, something useful may be a backtrace,
> but first, I would like to check whether it is not a problem related to
> ATLAS/Lapack,
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>

Hi David,
Thanks for the very prompt reply.

Below is the output you requested.

-gyro

-----
Here is the output of the tests:

$ python
Python 2.5.2 (r252:60911, Sep  5 2008, 07:14:57)
[GCC 4.1.2 20070626 (Red Hat 4.1.2-14)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import scipy
>>> scipy.test()
Running unit tests for scipy
NumPy version 1.2.0.dev5741
NumPy is installed in /home/gf/local/lib/python2.5/site-packages/numpy
SciPy version 0.7.0.dev4692
SciPy is installed in /home/gf/local/lib/python2.5/site-packages/scipy
Python version 2.5.2 (r252:60911, Sep  5 2008, 07:14:57) [GCC 4.1.2
20070626 (Red Hat 4.1.2-14)]
nose version 0.10.3

... [snip: a bunch of test results] ...

 warnings.warn(str1, DeprecationWarning)
.........E...........................FF.............ATLAS version
3.8.2 built by gf on Fri Sep  5 11:20:02 MDT 2008:
   UNAME    : Linux  2.6.18-53.el5xen #1 SMP Mon Nov 12 02:46:57 EST
2007 x86_64 x86_64 x86_64 GNU/Linux
   INSTFLG  : -1 0 -a 1
   ARCHDEFS : -DATL_OS_Linux -DATL_ARCH_PIII -DATL_CPUMHZ=2493
-DATL_SSE3 -DATL_SSE2 -DATL_SSE1 -DATL_USE64BITS -DATL_GAS_x8664
   F2CDEFS  : -DAdd_ -DF77_INTEGER=int -DStringSunStyle
   CACHEEDGE: 0
   F77      : gfortran, version GNU Fortran (GCC) 4.1.2 20070626 (Red
Hat 4.1.2-14)
   F77FLAGS : -fomit-frame-pointer -mfpmath=387 -O2 -falign-loops=4 -fPIC -m64
   SMC      : gcc, version gcc (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14)
   SMCFLAGS : -fomit-frame-pointer -mfpmath=387 -O2 -falign-loops=4 -fPIC -m64
   SKC      : gcc, version gcc (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14)
   SKCFLAGS : -fomit-frame-pointer -mfpmath=387 -O2 -falign-loops=4 -fPIC -m64
.............................................F.Segmentation fault

-----

Here is the top of the output from
$ python setup.py build

mkl_info:
  libraries mkl,vml,guide not found in /home/gf/local/lib/
  libraries mkl,vml,guide not found in /home/gf/local/lib64/
  libraries mkl,vml,guide not found in /home/gf/local/lib/atlas
  libraries mkl,vml,guide not found in /usr/local/lib
  NOT AVAILABLE

fftw3_info:
  libraries fftw3 not found in /home/gf/local/lib/
  libraries fftw3 not found in /home/gf/local/lib64/
  libraries fftw3 not found in /home/gf/local/lib/atlas
  libraries fftw3 not found in /usr/local/lib
  fftw3 not found
  NOT AVAILABLE

fftw2_info:
  libraries rfftw,fftw not found in /home/gf/local/lib/
  libraries rfftw,fftw not found in /home/gf/local/lib64/
  libraries rfftw,fftw not found in /home/gf/local/lib/atlas
  libraries rfftw,fftw not found in /usr/local/lib
  fftw2 not found
  NOT AVAILABLE

dfftw_info:
  libraries drfftw,dfftw not found in /home/gf/local/lib/
  libraries drfftw,dfftw not found in /home/gf/local/lib64/
  libraries drfftw,dfftw not found in /home/gf/local/lib/atlas
  libraries drfftw,dfftw not found in /usr/local/lib
  dfftw not found
  NOT AVAILABLE

djbfft_info:
  NOT AVAILABLE

blas_opt_info:
blas_mkl_info:
  libraries mkl,vml,guide not found in /home/gf/local/lib/
  libraries mkl,vml,guide not found in /home/gf/local/lib64/
  libraries mkl,vml,guide not found in /home/gf/local/lib/atlas
  libraries mkl,vml,guide not found in /usr/local/lib
  NOT AVAILABLE

atlas_blas_threads_info:
Setting PTATLASATLAS
Setting PTATLASATLAS
Setting PTATLASATLAS
  FOUND:
    libraries  ['ptf77blas', 'ptcblas', 'atlas']
    library_dirs  ['/home/gf/local/lib/atlas']
    language  c

customize GnuFCompiler
Found executable /usr/bin/g77
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler using config
compiling '_configtest.c':

/* This file is generated from numpy/distutils/system_info.py */
void ATL_buildinfo(void);
int main(void) {
  ATL_buildinfo();
  return 0;
}
C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O3
-Wall -Wstrict-prototypes -fPIC

compile options: '-c'
gcc: _configtest.c
gcc -pthread _configtest.o -L/home/gf/local/lib/atlas -lptf77blas
-lptcblas -latlas -o _configtest
ATLAS version 3.8.2 built by  on Fri Sep  5 11:20:02 MDT 2008:
   UNAME    : Linux  2.6.18-53.el5xen #1 SMP Mon Nov 12 02:46:57 EST
2007 x86_64 x86_64 x86_6
4 GNU/Linux
   INSTFLG  : -1 0 -a 1
   ARCHDEFS : -DATL_OS_Linux -DATL_ARCH_PIII -DATL_CPUMHZ2493
-DATL_SSE3 -DATL_SSE2 -DATL_SSE1 -DATL_USE64BITS -DATL_GA
S_x8664
   F2CDEFS  : -DAdd_ -DF77_INTEGERint -DStringSunStyle
   CACHEEDGE: 0
   F77      : gfortran, version GNU Fortran (GCC) 4.1.2 20070626 (Red
Hat 4.1.2-14)
   F77FLAGS : -fomit-frame-pointer -mfpmath387 -O2 -falign-loops4 -fPIC -m64
   SMC      : gcc, version gcc (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14)
   SMCFLAGS : -fomit-frame-pointer -mfpmath387 -O2 -falign-loops4 -fPIC -m64
   SKC      : gcc, version gcc (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14)
   SKCFLAGS : -fomit-frame-pointer -mfpmath387 -O2 -falign-loops4 -fPIC -m64
success!
removing: _configtest.c _configtest.o _configtest
  FOUND:
    libraries  ['ptf77blas', 'ptcblas', 'atlas']
    library_dirs  ['/home/gf/local/lib/atlas']
    language  c
    define_macros  [('ATLAS_INFO', '"\\"3.8.2\\""')]

ATLAS version 3.8.2
lapack_opt_info:
lapack_mkl_info:
  NOT AVAILABLE

atlas_threads_info:
Setting PTATLASATLAS
  libraries lapack_atlas not found in /home/gf/local/lib/atlas
numpy.distutils.system_info.atlas_threads_info
Setting PTATLASATLAS
Setting PTATLASATLAS
  FOUND:
    libraries  ['lapack', 'ptf77blas', 'ptcblas', 'atlas']
    library_dirs  ['/home/gf/local/lib/atlas']
    language  f77
	gnu: no Fortran 90 compiler found
	customize GnuFCompiler
	gnu: no Fortran 90 compiler found
	gnu: no Fortran 90 compiler found
	customize GnuFCompiler using config
	compiling '_configtest.c':

	/* This file is generated from numpy/distutils/system_info.py */
	void ATL_buildinfo(void);
	int main(void) {
	  ATL_buildinfo();
	  return 0;
	}
	C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O3
-Wall -Wstrict-prototypes -fPIC

	compile options: '-c'
	gcc: _configtest.c
	gcc -pthread _configtest.o -L/home/gf/local/lib/atlas -llapack
-lptf77blas -lptcblas -latlas -o _configtest
	ATLAS version 3.8.2 built by  on Fri Sep  5 11:20:02 MDT 2008:
	   UNAME    : Linux 2.6.18-53.el5xen #1 SMP Mon Nov 12 02:46:57 EST
2007 x86_64 x86_64 x86_6
	4 GNU/Linux
	   INSTFLG  : -1 0 -a 1
	   ARCHDEFS : -DATL_OS_Linux -DATL_ARCH_PIII -DATL_CPUMHZ2493
-DATL_SSE3 -DATL_SSE2 -DATL_SSE1 -DATL_USE64BITS -DATL_GA
	S_x8664
	   F2CDEFS  : -DAdd_ -DF77_INTEGERint -DStringSunStyle
	   CACHEEDGE: 0
	   F77      : gfortran, version GNU Fortran (GCC) 4.1.2 20070626 (Red
Hat 4.1.2-14)
	   F77FLAGS : -fomit-frame-pointer -mfpmath387 -O2 -falign-loops4 -fPIC -m64
	   SMC      : gcc, version gcc (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14)
	   SMCFLAGS : -fomit-frame-pointer -mfpmath387 -O2 -falign-loops4 -fPIC -m64
	   SKC      : gcc, version gcc (GCC) 4.1.2 20070626 (Red Hat 4.1.2-14)
	   SKCFLAGS : -fomit-frame-pointer -mfpmath387 -O2 -falign-loops4 -fPIC -m64
	success!
	removing: _configtest.c _configtest.o _configtest
	  FOUND:
	    libraries  ['lapack', 'ptf77blas', 'ptcblas', 'atlas']
	    library_dirs  ['/home/gf/local/lib/atlas']
	    language  f77
	    define_macros  [('ATLAS_INFO', '"\\"3.8.2\\""')]

	ATLAS version 3.8.2
	ATLAS version 3.8.2
	umfpack_info:
	  libraries umfpack not found in /home/gf/local/lib/
	  libraries umfpack not found in /home/gf/local/lib64/
	  libraries umfpack not found in /home/gf/local/lib/atlas
	  libraries umfpack not found in /usr/local/lib
	  NOT AVAILABLE


From david at ar.media.kyoto-u.ac.jp  Sun Sep  7 12:04:18 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 08 Sep 2008 01:04:18 +0900
Subject: [SciPy-user] segmentation fault in scipy.test()
In-Reply-To: <ace26be40809070915k52b11fdfuef092c6a2483a333@mail.gmail.com>
References: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>	<48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>
	<ace26be40809070915k52b11fdfuef092c6a2483a333@mail.gmail.com>
Message-ID: <48C3FB82.7040603@ar.media.kyoto-u.ac.jp>

gfunch wrote:
> Running unit tests for scipy
> NumPy version 1.2.0.dev5741
> NumPy is installed in /home/gf/local/lib/python2.5/site-packages/numpy
> SciPy version 0.7.0.dev4692
> SciPy is installed in /home/gf/local/lib/python2.5/site-packages/scipy
> Python version 2.5.2 (r252:60911, Sep  5 2008, 07:14:57) [GCC 4.1.2
> 20070626 (Red Hat 4.1.2-14)]
> nose version 0.10.3
>
> ... [snip: a bunch of test results] ...
>
>   

Sorry, I forgot to tell you to run the test in verbose mode
(scipy.test(verbose = 10)), because otherwise, the output is useless for
debugging purpose.

David


From gyromagnetic at gmail.com  Sun Sep  7 12:30:42 2008
From: gyromagnetic at gmail.com (gfunch)
Date: Sun, 7 Sep 2008 10:30:42 -0600
Subject: [SciPy-user] segmentation fault in scipy.test()
In-Reply-To: <48C3FB82.7040603@ar.media.kyoto-u.ac.jp>
References: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>
	<48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>
	<ace26be40809070915k52b11fdfuef092c6a2483a333@mail.gmail.com>
	<48C3FB82.7040603@ar.media.kyoto-u.ac.jp>
Message-ID: <ace26be40809070930l40b47073pc77f53048395026@mail.gmail.com>

On Sun, Sep 7, 2008 at 10:04 AM, David Cournapeau
<david at ar.media.kyoto-u.ac.jp> wrote:
>
> Sorry, I forgot to tell you to run the test in verbose mode
> (scipy.test(verbose = 10)), because otherwise, the output is useless for
> debugging purpose.
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>

Hi David,
Below are the 'non-ok' results (and context) from the output of
scipy.test(verbose=10).

-gyro

test_recasts (test_recaster.TestRecaster) ... ok
test_smallest_int_sctype (test_recaster.TestRecaster) ... ok
Failure: ImportError
(/home/breisfel/local/lib/python2.5/site-packages/scipy/lib/blas/fblas.so:
undefined symbol: _gfortran_st_write_done) ... ERROR
test_lapack.test_all_lapack ... ok
test_lapack.test_all_lapack ... ok

test_lapack.test_all_lapack ... ok
test_lapack.test_all_lapack ... ok
test_lapack.test_all_lapack ... FAILtest_lapack.test_all_lapack ... FAIL
test_lapack.test_all_lapack ... ok
test_lapack.test_all_lapack ... ok

test_fblas (test_blas.TestBLAS) ... ok
test_axpy (test_blas.TestCBLAS1Simple) ... ok
test_amax (test_blas.TestFBLAS1Simple) ... ok
test_asum (test_blas.TestFBLAS1Simple) ... FAIL
test_axpy (test_blas.TestFBLAS1Simple) ... ok
test_complex_dotc (test_blas.TestFBLAS1Simple) ... Segmentation fault


From david at ar.media.kyoto-u.ac.jp  Sun Sep  7 12:17:54 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 08 Sep 2008 01:17:54 +0900
Subject: [SciPy-user] segmentation fault in scipy.test()
In-Reply-To: <ace26be40809070930l40b47073pc77f53048395026@mail.gmail.com>
References: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>	<48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>	<ace26be40809070915k52b11fdfuef092c6a2483a333@mail.gmail.com>	<48C3FB82.7040603@ar.media.kyoto-u.ac.jp>
	<ace26be40809070930l40b47073pc77f53048395026@mail.gmail.com>
Message-ID: <48C3FEB2.3060601@ar.media.kyoto-u.ac.jp>

gfunch wrote:
>
> Hi David,
> Below are the 'non-ok' results (and context) from the output of
> scipy.test(verbose=10).
>   

As I suspected, a fortran/atlas issue. I guess you have both g77 and
gfortran, than you built atlas with gfortran for the F77 interface, and
used g77 to build scipy. This cannot work. You should rebuild both numpy
and scipy from scratch (delete both installation directories AND build
directories), and use gfortran:

python setup.py build --fcompiler=gnu95 ...

cheers,

David


From gyromagnetic at gmail.com  Sun Sep  7 12:53:37 2008
From: gyromagnetic at gmail.com (gfunch)
Date: Sun, 7 Sep 2008 10:53:37 -0600
Subject: [SciPy-user] segmentation fault in scipy.test()
In-Reply-To: <48C3FEB2.3060601@ar.media.kyoto-u.ac.jp>
References: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>
	<48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>
	<ace26be40809070915k52b11fdfuef092c6a2483a333@mail.gmail.com>
	<48C3FB82.7040603@ar.media.kyoto-u.ac.jp>
	<ace26be40809070930l40b47073pc77f53048395026@mail.gmail.com>
	<48C3FEB2.3060601@ar.media.kyoto-u.ac.jp>
Message-ID: <ace26be40809070953rec283d1ta8a70648e0a4817c@mail.gmail.com>

On Sun, Sep 7, 2008 at 10:17 AM, David Cournapeau
<david at ar.media.kyoto-u.ac.jp> wrote:
> gfunch wrote:
>>
>> Hi David,
>> Below are the 'non-ok' results (and context) from the output of
>> scipy.test(verbose=10).
>>
>
> As I suspected, a fortran/atlas issue. I guess you have both g77 and
> gfortran, than you built atlas with gfortran for the F77 interface, and
> used g77 to build scipy. This cannot work. You should rebuild both numpy
> and scipy from scratch (delete both installation directories AND build
> directories), and use gfortran:
>
> python setup.py build --fcompiler=gnu95 ...
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>

Hi David,
Yes, that's it! Thanks.

I do get a few errors and failures:
ERROR: Failure: AttributeError ('module' object has no attribute
'knownfailureif')

FAIL: test_lapack.test_all_lapack
AssertionError:
Arrays are not almost equal

FAILED (SKIP=14, errors=3, failures=2)

Is this 'normal'?

Thanks, again!

-gyro


From david at ar.media.kyoto-u.ac.jp  Sun Sep  7 12:48:45 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 08 Sep 2008 01:48:45 +0900
Subject: [SciPy-user] segmentation fault in scipy.test()
In-Reply-To: <ace26be40809070953rec283d1ta8a70648e0a4817c@mail.gmail.com>
References: <ace26be40809070837u9b3a29emd7f83f1919340c79@mail.gmail.com>	<48C3F2E9.4050906@ar.media.kyoto-u.ac.jp>	<ace26be40809070915k52b11fdfuef092c6a2483a333@mail.gmail.com>	<48C3FB82.7040603@ar.media.kyoto-u.ac.jp>	<ace26be40809070930l40b47073pc77f53048395026@mail.gmail.com>	<48C3FEB2.3060601@ar.media.kyoto-u.ac.jp>
	<ace26be40809070953rec283d1ta8a70648e0a4817c@mail.gmail.com>
Message-ID: <48C405ED.8000006@ar.media.kyoto-u.ac.jp>

gfunch wrote:
> Hi David,
> Yes, that's it! Thanks.
>   

Cool

> I do get a few errors and failures:
> ERROR: Failure: AttributeError ('module' object has no attribute
> 'knownfailureif')
>   

You should update your numpy to a more recent version: those are really
recent (a couple of hours ago) changes for the soon to be released numpy
1.2, which next scipy release will depend on.

Those are harmless, though, if you don't want to waste time on updating
numpy,

cheers,

David


From david.strozzi at gmail.com  Sun Sep  7 22:18:56 2008
From: david.strozzi at gmail.com (David Strozzi)
Date: Sun, 7 Sep 2008 19:18:56 -0700
Subject: [SciPy-user] modified bessel function of the 2nd/3rd kind
Message-ID: <f21763420809071918p6543d48fvde176e24fa6d4038@mail.gmail.com>

Folks,

scipy.special contains these Bessel functions:

Bessel Functions
----------------

*  jn       -- Bessel function of integer order and real argument.
*  jv       -- Bessel function of real-valued order and complex argument.
*  jve      -- Exponentially scaled Bessel function.
*  yn       -- Bessel function of second kind (integer order).
*  yv       -- Bessel function of the second kind (real-valued order).
*  yve      -- Exponentially scaled Bessel function of the second kind.
*  kn       -- Modified Bessel function of the third kind (integer order).
*  kv       -- Modified Bessel function of the third kind (real order).
*  kve      -- Exponentially scaled modified Bessel function of the third kind.

As has been noted on this list before, kn, kv are also, in fact much
more commonly, referred to as mod. Bess. funcs of the *2nd* kind.  Can
this info please be updated?  Is there any good reason not to have the
help message say "of the second kind (sometimes called the third
kind)"?

As someone who works with these I have never heard them called the 3rd
kind, only the 2nd.

Thanks,
Dave Strozzi


From bernardo.rocha at meduni-graz.at  Mon Sep  8 11:03:49 2008
From: bernardo.rocha at meduni-graz.at (bernardo martins rocha)
Date: Mon, 08 Sep 2008 17:03:49 +0200
Subject: [SciPy-user] Illegal Instruction
Message-ID: <48C53ED5.3060905@meduni-graz.at>

Hi Guys,

i'm trying to use scipy.optimize.leastsq but whenever I call it I get 
the following error:

/"In [13]: p0 = [0.5,1,0.5,1] # initial guesses

In [14]: guessfit = dbexpl(t,p0)

In [15]: pbest = leastsq(residuals,p0,args=(data,t),full_output=1)
Illegal instruction"

/What does this error mean? I don't think it's is a error in the way I'm 
calling the method because I haven't changed my script that was working 
fine some weeks ago.

Thanks in advance!
Bernardo M. Rocha
/
/



From robert.kern at gmail.com  Mon Sep  8 11:47:14 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 8 Sep 2008 10:47:14 -0500
Subject: [SciPy-user] Illegal Instruction
In-Reply-To: <48C53ED5.3060905@meduni-graz.at>
References: <48C53ED5.3060905@meduni-graz.at>
Message-ID: <3d375d730809080847j1b0ef3f7ydf26f811784143a6@mail.gmail.com>

On Mon, Sep 8, 2008 at 10:03, bernardo martins rocha
<bernardo.rocha at meduni-graz.at> wrote:
> Hi Guys,
>
> i'm trying to use scipy.optimize.leastsq but whenever I call it I get
> the following error:
>
> /"In [13]: p0 = [0.5,1,0.5,1] # initial guesses
>
> In [14]: guessfit = dbexpl(t,p0)
>
> In [15]: pbest = leastsq(residuals,p0,args=(data,t),full_output=1)
> Illegal instruction"
>
> /What does this error mean? I don't think it's is a error in the way I'm
> calling the method because I haven't changed my script that was working
> fine some weeks ago.

Are you on Windows? Which scipy binaries are you using? Does your CPU
support SSE2?

-- 
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 contact at pythonxy.com  Mon Sep  8 14:01:29 2008
From: contact at pythonxy.com (Pierre Raybaut)
Date: Mon, 08 Sep 2008 20:01:29 +0200
Subject: [SciPy-user] [ Python(x,y) ] New release : 2.0.5
Message-ID: <48C56879.7070607@pythonxy.com>

Hi all,

As you may already know, Python(x,y) is a free scientific-oriented
Python Distribution based on Qt and Eclipse providing a self-consistent
scientific development environment.

Release 2.0.5 is now available on http://www.pythonxy.com.
(Full Edition, Basic Edition, Light Edition and Update)

Changes history
Version 2.0.5 (09-06-2008)

     * Updated:
           o Enthought Tool Suite 3.0.0(.1) (docs and examples updated)
           o PyDAP 2.2.6.5
           o xy 1.0.5 (New shortcuts and help links)
     * Corrected:
           o VTK: VTKData folder was not found by the example scripts
(see VTK documentation folder)
           o Eclipse/Windows Vista: Java RE updated (version 6 Update 7)
in Eclipse main installer - the previous version was freezing on some
machines under Windows Vista
           o Console 2 (New settings)
           o Notepad++ (New Console 2 settings)

Regards,
Pierre Raybaut




From pav at iki.fi  Mon Sep  8 14:59:05 2008
From: pav at iki.fi (Pauli Virtanen)
Date: Mon, 8 Sep 2008 18:59:05 +0000 (UTC)
Subject: [SciPy-user] modified bessel function of the 2nd/3rd kind
References: <f21763420809071918p6543d48fvde176e24fa6d4038@mail.gmail.com>
Message-ID: <ga3slo$87j$3@ger.gmane.org>

Sun, 07 Sep 2008 19:18:56 -0700, David Strozzi wrote:
> Folks,
> 
> scipy.special contains these Bessel functions:
> 
> Bessel Functions
> ----------------
[clip]
> 
> As has been noted on this list before, kn, kv are also, in fact much
> more commonly, referred to as mod. Bess. funcs of the *2nd* kind.  Can
> this info please be updated?  Is there any good reason not to have the
> help message say "of the second kind (sometimes called the third kind)"?
> 
> As someone who works with these I have never heard them called the 3rd
> kind, only the 2nd.

Changed in r4704; "of the third kind" seems to be some kind of historical 
remnant that's rarely used today.

-- 
Pauli Virtanen



From bernardo.rocha at meduni-graz.at  Mon Sep  8 15:59:50 2008
From: bernardo.rocha at meduni-graz.at (Bernardo Martins Rocha)
Date: Mon, 08 Sep 2008 21:59:50 +0200
Subject: [SciPy-user] Illegal instruction
Message-ID: <48C5A0560200002A0001C194@si062.meduni-graz.at>

Hi..

I running Opensuse 11 in an AMD Opteron x64. So, I think the CPU
supports SSE2. I'm using the scipy binaries provided by the
opensuse-science repository. The strange thing that I also tried some
examples (written by Travis Oliphant) and I?got the same error when I
call leastsq.

I did the scipy.test(level=1) and I got this error too. 

Should I try to install the scipy from the svn repository?

Thanks!
Best regards,
Bernardo M. Rocha

Message: 3
Date: Mon, 08 Sep 2008 17:03:49 +0200
From: bernardo martins rocha <bernardo.rocha at meduni-graz.at>
Subject: [SciPy-user] Illegal Instruction
To: scipy-user at scipy.org
Message-ID: <48C53ED5.3060905 at meduni-graz.at>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Hi Guys,

i'm trying to use scipy.optimize.leastsq but whenever I call it I get 
the following error:

/"In [13]: p0 = [0.5,1,0.5,1] # initial guesses

In [14]: guessfit = dbexpl(t,p0)

In [15]: pbest = leastsq(residuals,p0,args=(data,t),full_output=1)
Illegal instruction"

/What does this error mean? I don't think it's is a error in the way I'm

calling the method because I haven't changed my script that was working 
fine some weeks ago.

Thanks in advance!
Bernardo M. Rocha
/
/



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

Message: 4
Date: Mon, 8 Sep 2008 10:47:14 -0500
From: "Robert Kern" <robert.kern at gmail.com>
Subject: Re: [SciPy-user] Illegal Instruction
To: "SciPy Users List" <scipy-user at scipy.org>
Message-ID:
	<3d375d730809080847j1b0ef3f7ydf26f811784143a6 at mail.gmail.com>
Content-Type: text/plain; charset=UTF-8

On Mon, Sep 8, 2008 at 10:03, bernardo martins rocha
<bernardo.rocha at meduni-graz.at> wrote:
> Hi Guys,
>
> i'm trying to use scipy.optimize.leastsq but whenever I call it I get
> the following error:
>
> /"In [13]: p0 = [0.5,1,0.5,1] # initial guesses
>
> In [14]: guessfit = dbexpl(t,p0)
>
> In [15]: pbest = leastsq(residuals,p0,args=(data,t),full_output=1)
> Illegal instruction"
>
> /What does this error mean? I don't think it's is a error in the way
I'm
> calling the method because I haven't changed my script that was
working
> fine some weeks ago.

Are you on Windows? Which scipy binaries are you using? Does your CPU
support SSE2?

-- 
Robert Kern

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


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

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


End of SciPy-user Digest, Vol 61, Issue 10
******************************************



From FDU.xiaojf at gmail.com  Tue Sep  9 04:47:47 2008
From: FDU.xiaojf at gmail.com (xiaojf)
Date: Tue, 9 Sep 2008 01:47:47 -0700 (PDT)
Subject: [SciPy-user] maybe a bug in scipy.io.loadmat
Message-ID: <e39f1f51-d40e-46ae-b07b-ab50bf8d644d@r66g2000hsg.googlegroups.com>

Hi all,

I found that scipy.io.loadmat(filename) doesn't work when filename
contains '/' instead of '\\' as file path separater.

"""
Help on function loadmat in module scipy.io.mio:

loadmat(file_name, mdict=None, appendmat=True, basename='raw',
**kwargs)
    Load Matlab(tm) file

    file_name          - Name of the mat file
                         (do not need .mat extension if
appendmat==True)
                         If name not a full path name, search for the
file on
                         the sys.path list and use the first one found
(the
                         current directory is searched first).
                         Can also pass open file-like object
"""

The problem is in function find_mat_file, which tries to find the file
in directories listed in sys.path when there is no os.sep in
file_name.

I couldn't understand why the function find_mat_file() tries to
find .mat file in directories listed in sys.path, since sys.path is
module search path not data file search path.

def find_mat_file(file_name, appendmat=True):
    ''' Try to find .mat file on system path

    file_name     - file name string
    append_mat    - If True, and file_name does not end in '.mat',
appends it
    '''
    if appendmat and file_name[-4:] == ".mat":
        file_name = file_name[:-4]
    if os.sep in file_name:
        full_name = file_name
        if appendmat:
            full_name = file_name + ".mat"
    else:
        full_name = None
        junk, file_name = os.path.split(file_name)
        for path in sys.path:
            test_name = os.path.join(path, file_name)
            if appendmat:
                test_name += ".mat"
            try:
                fid = open(test_name,'rb')
                fid.close()
                full_name = test_name
                break
            except IOError:
                pass
    return full_name


From cyril.giraudon at free.fr  Tue Sep  9 09:37:26 2008
From: cyril.giraudon at free.fr (cyril giraudon)
Date: Tue, 09 Sep 2008 15:37:26 +0200
Subject: [SciPy-user] butterworth filter
Message-ID: <48C67C16.9030701@free.fr>

Hi,

I use scipy 0.6.0 and i try to reproduce the plot of the matlab butter 
function web documentation (first response for a google request "matlab 
butter example").

The matlab code is :

[z,p,k] = butter(9,300/500,'high');
[sos,g] = zp2sos(z,p,k);	     % Convert to SOS form
Hd = dfilt.df2tsos(sos,g);   % Create a dfilt object
h = fvtool(Hd);	             % Plot magnitude response
set(h,'Analysis','freq')	     % Display frequency response 


In scipy, I write :

from scipy.signal import butter, freqz

from pylab import show, grid, log, plot

b, a = butter(9, 300./500., 'high')

fi = freqz(b, a)

plot(fi[0], 20*log(abs(fi[1])))

grid()

show()



Why are the two filters not the same ?


Thanks a lot,

Cyril.




From kdere at gmu.edu  Tue Sep  9 12:12:00 2008
From: kdere at gmu.edu (Ken Dere)
Date: Tue, 9 Sep 2008 16:12:00 +0000 (UTC)
Subject: [SciPy-user] need an IDL-like rebin function
Message-ID: <loom.20080909T160916-520@post.gmane.org>

I have tried to use zoom as a rebin-like function but it's behavior at the 
edges of the array is not acceptable and I don't seem to be able to influence 
this behavior.  Actually, I don't even want interpolation at this point.

advice appreciated

Ken Dere



From gael.varoquaux at normalesup.org  Tue Sep  9 12:17:44 2008
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Tue, 9 Sep 2008 18:17:44 +0200
Subject: [SciPy-user] need an IDL-like rebin function
In-Reply-To: <loom.20080909T160916-520@post.gmane.org>
References: <loom.20080909T160916-520@post.gmane.org>
Message-ID: <20080909161744.GC28438@phare.normalesup.org>

On Tue, Sep 09, 2008 at 04:12:00PM +0000, Ken Dere wrote:
> I have tried to use zoom as a rebin-like function but it's behavior at the 
> edges of the array is not acceptable and I don't seem to be able to influence 
> this behavior.  Actually, I don't even want interpolation at this point.

You might find useful examples on the following page:

http://www.scipy.org/Cookbook/Rebinning

HTH,

Ga?l


From bryan at ideotrope.org  Tue Sep  9 12:48:14 2008
From: bryan at ideotrope.org (Bryan Keith)
Date: Tue, 9 Sep 2008 10:48:14 -0600 (MDT)
Subject: [SciPy-user] scipy.test fails: clapack module is empty
Message-ID: <4967.64.78.232.178.1220978894.squirrel@ideotrope.org>

Hello,

I'm trying to install scipy (0.6.0) on Ubuntu 8.04 64 bit with Python
2.5.2 and numpy 1.0.4

Installation (via apt-get) seems to go fine, but the test suite fails. 
I'm not sure how concerned I should be about this failure.  I've searched
for some of the errors that I'm getting but couldn't find anything to
either resolve the tests and make me decide to ignore them.  I'm pasting
the test results below.  Any help is appreciated.  Thank you.

Bryan

>>> import scipy
>>> scipy.test(level=1)
  Found 9/9 tests for scipy.cluster.vq
  Found 18/18 tests for scipy.fftpack.basic
  Found 4/4 tests for scipy.fftpack.helper
  Found 20/20 tests for scipy.fftpack.pseudo_diffs
  Found 1/1 tests for scipy.integrate
  Found 10/10 tests for scipy.integrate.quadpack
  Found 3/3 tests for scipy.integrate.quadrature
  Found 6/6 tests for scipy.interpolate
  Found 6/6 tests for scipy.interpolate.fitpack
  Found 4/4 tests for scipy.io.array_import
  Found 28/28 tests for scipy.io.mio
  Found 13/13 tests for scipy.io.mmio
  Found 5/5 tests for scipy.io.npfile
  Found 4/4 tests for scipy.io.recaster
  Found 16/16 tests for scipy.lib.blas
  Found 128/128 tests for scipy.lib.blas.fblas

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

  Found 42/42 tests for scipy.lib.lapack
  Found 41/41 tests for scipy.linalg.basic
<module 'scipy.linalg.fblas' from
'/usr/lib/python2.5/site-packages/scipy/linalg/fblas.so'>
  Found 16/16 tests for scipy.linalg.blas
  Found 72/72 tests for scipy.linalg.decomp
  Found 128/128 tests for scipy.linalg.fblas
  Found 6/6 tests for scipy.linalg.iterative
  Found 4/4 tests for scipy.linalg.lapack
  Found 7/7 tests for scipy.linalg.matfuncs
  Found 9/9 tests for scipy.linsolve.umfpack
  Found 2/2 tests for scipy.maxentropy
  Found 3/3 tests for scipy.misc.pilutil
  Found 399/399 tests for scipy.ndimage
  Found 5/5 tests for scipy.odr
  Found 8/8 tests for scipy.optimize
  Found 1/1 tests for scipy.optimize.cobyla
  Found 10/10 tests for scipy.optimize.nonlin
  Found 4/4 tests for scipy.optimize.zeros
  Found 5/5 tests for scipy.signal.signaltools
  Found 4/4 tests for scipy.signal.wavelets
  Found 152/152 tests for scipy.sparse
  Found 342/342 tests for scipy.special.basic
  Found 3/3 tests for scipy.special.spfun_stats
  Found 107/107 tests for scipy.stats
  Found 73/73 tests for scipy.stats.distributions
  Found 10/10 tests for scipy.stats.morestats
  Found 0/0 tests for __main__
.../usr/lib/python2.5/site-packages/scipy/cluster/vq.py:477: UserWarning:
One of the clusters is empty. Re-run kmean with a different
initialization.
  warnings.warn("One of the clusters is empty. "
exception raised as expected: One of the clusters is empty. Re-run kmean
with a different initialization.
................................................Residual: 1.05006987327e-07
..................../usr/lib/python2.5/site-packages/scipy/interpolate/fitpack2.py:458:
UserWarning:
The coefficients of the spline returned have been computed as the
minimal norm least-squares solution of a (numerically) rank deficient
system (deficiency=7). If deficiency is large, the results may be
inaccurate. Deficiency may strongly depend on the value of eps.
  warnings.warn(message)
......
Don't worry about a warning regarding the number of bytes read.
Warning: 1000000 bytes requested, 20 bytes read.
.........................................................................caxpy:n=4
..caxpy:n=3
....ccopy:n=4
..ccopy:n=3
.............cscal:n=4
....cswap:n=4
..cswap:n=3
.....daxpy:n=4
..daxpy:n=3
....dcopy:n=4
..dcopy:n=3
.............dscal:n=4
....dswap:n=4
..dswap:n=3
.....saxpy:n=4
..saxpy:n=3
....scopy:n=4
..scopy:n=3
.............sscal:n=4
....sswap:n=4
..sswap:n=3
.....zaxpy:n=4
..zaxpy:n=3
....zcopy:n=4
..zcopy:n=3
.............zscal:n=4
....zswap:n=4
..zswap:n=3
............................FF.......................................................
****************************************************************
WARNING: cblas module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses fblas instead of cblas.
****************************************************************

...........................................................................................caxpy:n=4
..caxpy:n=3
....ccopy:n=4
..ccopy:n=3
.............cscal:n=4
....cswap:n=4
..cswap:n=3
.....daxpy:n=4
..daxpy:n=3
....dcopy:n=4
..dcopy:n=3
.............dscal:n=4
....dswap:n=4
..dswap:n=3
.....saxpy:n=4
..saxpy:n=3
....scopy:n=4
..scopy:n=3
.............sscal:n=4
....sswap:n=4
..sswap:n=3
.....zaxpy:n=4
..zaxpy:n=3
....zcopy:n=4
..zcopy:n=3
.............zscal:n=4
....zswap:n=4
..zswap:n=3
..........
****************************************************************
WARNING: clapack module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses flapack instead of clapack.
****************************************************************

...Result may be inaccurate, approximate err = 1.11929781998e-08
...Result may be inaccurate, approximate err = 7.73070496507e-12
......Use minimum degree ordering on A'+A.
..Use minimum degree ordering on A'+A.
F..Use minimum degree ordering on A'+A.
F............................................................................................................/usr/lib/python2.5/site-packages/scipy/ndimage/interpolation.py:41:
UserWarning: Mode "reflect" may yield incorrect results on boundaries.
Please use "mirror" instead.
  warnings.warn('Mode "reflect" may yield incorrect results on '
........................................................................................................................................................................................................................................................................................................F..F.........................................................Use
minimum degree ordering on A'+A.
.....................................Use minimum degree ordering on A'+A.
.....................................Use minimum degree ordering on A'+A.
................................Use minimum degree ordering on A'+A.
....................................................................................................................................................................................................................................................................................................................................................0.2
0.2
0.2
......0.2
..0.2
0.2
0.2
0.2
0.2
.........................................................................................................................................................................................................Ties
preclude use of exact statistic.
..Ties preclude use of exact statistic.
......
======================================================================
FAIL: check_syevr (scipy.lib.tests.test_lapack.test_flapack_float)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py",
line 41, in check_syevr
    assert_array_almost_equal(w,exact_w)
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769474,  9.18222618], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])

======================================================================
FAIL: check_syevr_irange (scipy.lib.tests.test_lapack.test_flapack_float)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py",
line 66, in check_syevr_irange
    assert_array_almost_equal(w,exact_w[rslice])
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769474,  9.18222618], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])

======================================================================
FAIL: Solve: single precision complex
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/usr/lib/python2.5/site-packages/scipy/linsolve/umfpack/tests/test_umfpack.py",
line 32, in check_solve_complex_without_umfpack
    assert_array_almost_equal(a*x, b)
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 20.0%)
 x: array([ 1.00000000+0.j,  1.99999809+0.j,  3.00000000+0.j, 
4.00000048+0.j,
        5.00000000+0.j], dtype=complex64)
 y: array([1, 2, 3, 4, 5])

======================================================================
FAIL: Solve: single precision
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/usr/lib/python2.5/site-packages/scipy/linsolve/umfpack/tests/test_umfpack.py",
line 43, in check_solve_without_umfpack
    assert_array_almost_equal(a*x, b)
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 20.0%)
 x: array([ 1.        ,  1.99999809,  3.        ,  4.00000048,  5.       
], dtype=float32)
 y: array([1, 2, 3, 4, 5])

======================================================================
FAIL: test_explicit (scipy.tests.test_odr.test_odr)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
line 50, in test_explicit
    -8.7849712165253724e-02]),
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

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

======================================================================
FAIL: test_multi (scipy.tests.test_odr.test_odr)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.5/site-packages/scipy/odr/tests/test_odr.py",
line 191, in test_multi
    0.5101147161764654,  0.5173902330489161]),
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
232, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line
217, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

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

----------------------------------------------------------------------
Ran 1728 tests in 7.579s

FAILED (failures=6)
<unittest._TextTestResult run=1728 errors=0 failures=6>
>>>




From bernardo.rocha at meduni-graz.at  Tue Sep  9 13:02:46 2008
From: bernardo.rocha at meduni-graz.at (bernardo martins rocha)
Date: Tue, 09 Sep 2008 19:02:46 +0200
Subject: [SciPy-user] scipy.test fails: clapack module is empty (Bryan
	Keith)
In-Reply-To: <mailman.2013.1220978900.3376.scipy-user@scipy.org>
References: <mailman.2013.1220978900.3376.scipy-user@scipy.org>
Message-ID: <48C6AC36.6030808@meduni-graz.at>

Hi Bryan Keith,
>    8. scipy.test fails: clapack module is empty (Bryan Keith)
>   
I've got the same error with my installation (opensuse11 via yast --- 
science repository), please have a look at the message below. I also 
have a "Illegal Instruction" error in the end of the test, which is 
worse than your problem. I would like to get rid off this. Any suggestions?

Thanks!

--------------------------
In [1]: import scipy

In [2]: scipy.test()
Failed importing scipy.linsolve.umfpack: 'module' object has no 
attribute 'umfpack'
Found 9/9 tests for scipy.cluster.tests.test_vq
Found 20/20 tests for scipy.fftpack.tests.test_pseudo_diffs
Found 4/4 tests for scipy.fftpack.tests.test_helper
Found 18/18 tests for scipy.fftpack.tests.test_basic
Found 1/1 tests for scipy.integrate.tests.test_integrate
Found 3/3 tests for scipy.integrate.tests.test_quadrature
Found 10/10 tests for scipy.integrate.tests.test_quadpack
Found 6/6 tests for scipy.tests.test_fitpack
Found 6/6 tests for scipy.tests.test_interpolate
Found 28/28 tests for scipy.io.tests.test_mio
Found 4/4 tests for scipy.io.tests.test_recaster
Found 5/5 tests for scipy.io.tests.test_npfile
Found 13/13 tests for scipy.io.tests.test_mmio
Found 4/4 tests for scipy.io.tests.test_array_import
Found 16/16 tests for scipy.lib.blas.tests.test_blas
Found 128/128 tests for scipy.lib.blas.tests.test_fblas

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

Found 42/42 tests for scipy.lib.lapack.tests.test_lapack
Found 4/4 tests for scipy.linalg.tests.test_lapack
<module 'scipy.linalg.fblas' from 
'/usr/lib64/python2.5/site-packages/scipy/linalg/fblas.so'>
Found 16/16 tests for scipy.linalg.tests.test_blas
Found 6/6 tests for scipy.linalg.tests.test_iterative
Found 41/41 tests for scipy.linalg.tests.test_basic
Found 128/128 tests for scipy.linalg.tests.test_fblas
Found 7/7 tests for scipy.linalg.tests.test_matfuncs
Found 72/72 tests for scipy.linalg.tests.test_decomp
Failed importing 
/usr/lib64/python2.5/site-packages/scipy/linsolve/umfpack/tests/test_umfpack.py: 
'module' object has no attribute 'umfpack'
Found 2/2 tests for scipy.maxentropy.tests.test_maxentropy
Found 3/3 tests for scipy.misc.tests.test_pilutil
Found 399/399 tests for scipy.ndimage.tests.test_ndimage
Found 5/5 tests for scipy.odr.tests.test_odr
Found 4/4 tests for scipy.optimize.tests.test_zeros
Found 1/1 tests for scipy.optimize.tests.test_cobyla
Found 8/8 tests for scipy.optimize.tests.test_optimize
Found 10/10 tests for scipy.optimize.tests.test_nonlin
Found 5/5 tests for scipy.signal.tests.test_signaltools
Found 4/4 tests for scipy.signal.tests.test_wavelets
Found 152/152 tests for scipy.sparse.tests.test_sparse
Found 3/3 tests for scipy.special.tests.test_spfun_stats
Found 342/342 tests for scipy.special.tests.test_basic
Found 10/10 tests for scipy.stats.tests.test_morestats
Found 107/107 tests for scipy.stats.tests.test_stats
Found 73/73 tests for scipy.stats.tests.test_distributions
Found 0/0 tests for scipy.weave.tests.test_c_spec
Found 1/1 tests for scipy.weave.tests.test_ast_tools
Found 9/9 tests for scipy.weave.tests.test_build_tools
Found 2/2 tests for scipy.weave.tests.test_blitz_tools
building extensions here: /home/rocha/.python25_compiled/m11
Found 1/1 tests for scipy.weave.tests.test_ext_tools
Found 0/0 tests for scipy.weave.tests.test_inline_tools
Found 26/26 tests for scipy.weave.tests.test_catalog
Found 0/0 tests for scipy.weave.tests.test_scxx_sequence
Found 74/74 tests for scipy.weave.tests.test_size_check
Failed importing 
/usr/lib64/python2.5/site-packages/scipy/weave/tests/old_test_wx_spec.py: 
Could not locate wxPython base directory.
Found 0/0 tests for scipy.weave.tests.test_scxx_dict
Found 0/0 tests for scipy.weave.tests.test_scxx_object
Found 3/3 tests for scipy.weave.tests.test_standard_array_spec
Found 16/16 tests for scipy.weave.tests.test_slice_handler
Failed importing 
/usr/lib64/python2.5/site-packages/scipy/weave/tests/test_wx_spec.py: 
Could not locate wxPython base directory.
../usr/lib64/python2.5/site-packages/scipy/cluster/vq.py:477: 
UserWarning: One of the clusters is empty. Re-run kmean with a different 
initialization.
warnings.warn("One of the clusters is empty. "
exception raised as expected: One of the clusters is empty. Re-run kmean 
with a different initialization.
..........Illegal instruction

Bernardo M. Rocha




From kdere at gmu.edu  Tue Sep  9 14:05:50 2008
From: kdere at gmu.edu (Ken Dere)
Date: Tue, 9 Sep 2008 18:05:50 +0000 (UTC)
Subject: [SciPy-user] need an IDL-like rebin function
References: <loom.20080909T160916-520@post.gmane.org>
	<20080909161744.GC28438@phare.normalesup.org>
Message-ID: <loom.20080909T180415-677@post.gmane.org>

Gael Varoquaux <gael.varoquaux <at> normalesup.org> writes:

> 
> On Tue, Sep 09, 2008 at 04:12:00PM +0000, Ken Dere wrote:
> > I have tried to use zoom as a rebin-like function but it's behavior at the 
> > edges of the array is not acceptable and I don't seem to be able to 
influence 
> > this behavior.  Actually, I don't even want interpolation at this point.
> 
> You might find useful examples on the following page:
> 
> http://www.scipy.org/Cookbook/Rebinning
> 
> HTH,
> 
> Ga?l
> 

Thanks.  I looked at them.  I could not get them to work and their indexing 
logic was so complicated I would never be able to fix them.

Ken





From amcmorl at gmail.com  Tue Sep  9 14:57:05 2008
From: amcmorl at gmail.com (Angus McMorland)
Date: Tue, 9 Sep 2008 14:57:05 -0400
Subject: [SciPy-user] need an IDL-like rebin function
In-Reply-To: <loom.20080909T180415-677@post.gmane.org>
References: <loom.20080909T160916-520@post.gmane.org>
	<20080909161744.GC28438@phare.normalesup.org>
	<loom.20080909T180415-677@post.gmane.org>
Message-ID: <e85f13c40809091157x140f000i7cefa1abcb38213c@mail.gmail.com>

2008/9/9 Ken Dere <kdere at gmu.edu>:
> Gael Varoquaux <gael.varoquaux <at> normalesup.org> writes:
>
>>
>> On Tue, Sep 09, 2008 at 04:12:00PM +0000, Ken Dere wrote:
>> > I have tried to use zoom as a rebin-like function but it's behavior at the
>> > edges of the array is not acceptable and I don't seem to be able to
> influence
>> > this behavior.  Actually, I don't even want interpolation at this point.
>>
>> You might find useful examples on the following page:
>>
>> http://www.scipy.org/Cookbook/Rebinning
>>
>> HTH,
>>
>> Ga?l
>>
>
> Thanks.  I looked at them.  I could not get them to work and their indexing
> logic was so complicated I would never be able to fix them.

What wasn't working? It's good for the cookbook examples to be a)
functional and b) well documented, so if there is a problem we should
try to fix it. All the examples have worked correctly for me in the
past, but I haven't used them for a while.

Angus.
-- 
AJC McMorland
Post-doctoral research fellow
Neurobiology, University of Pittsburgh


From teddy.kord at googlemail.com  Tue Sep  9 15:54:02 2008
From: teddy.kord at googlemail.com (Ish Aden)
Date: Tue, 9 Sep 2008 20:54:02 +0100
Subject: [SciPy-user] Pytrilinos on Windows
Message-ID: <34d365e70809091254x7ec84543y13b032aae141cd6b@mail.gmail.com>

Hello

Could anyone who's successfully installed Pytrilinos on a Windows machine
explain how they did it.

Also, are there any mature/close to mature Python PDE solvers out there in
addition to 'sfepy'.

Thanks in advance.

Ted
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080909/84cd9e09/attachment.html>

From bgoli at sun.ac.za  Tue Sep  9 16:12:29 2008
From: bgoli at sun.ac.za (Brett G. Olivier)
Date: Tue, 09 Sep 2008 22:12:29 +0200
Subject: [SciPy-user] Pytrilinos on Windows
In-Reply-To: <34d365e70809091254x7ec84543y13b032aae141cd6b@mail.gmail.com>
References: <34d365e70809091254x7ec84543y13b032aae141cd6b@mail.gmail.com>
Message-ID: <48C6D8AD.4080608@sun.ac.za>

Ish Aden wrote:
> Could anyone who's successfully installed Pytrilinos on a Windows 
> machine explain how they did it.
> 
> Also, are there any mature/close to mature Python PDE solvers out there 
> in addition to 'sfepy'.
Have you had a look at FiPy (http://www.ctcms.nist.gov/fipy/)

Cheers
Brett


From kpdere at verizon.net  Tue Sep  9 17:25:22 2008
From: kpdere at verizon.net (Ken Dere)
Date: Tue, 9 Sep 2008 21:25:22 +0000 (UTC)
Subject: [SciPy-user] need an IDL-like rebin function - operator error
References: <loom.20080909T160916-520@post.gmane.org>
Message-ID: <loom.20080909T212410-694@post.gmane.org>

Ken Dere <kdere <at> gmu.edu> writes:

> 
> I have tried to use zoom as a rebin-like function but it's behavior at the 
> edges of the array is not acceptable and I don't seem to be able to influence 
> this behavior.  Actually, I don't even want interpolation at this point.
> 
> advice appreciated
> 
> Ken Dere
> 

My mistake.  It really does work.

Sorry for the bother.

Ken






From teddy.kord at googlemail.com  Tue Sep  9 19:28:21 2008
From: teddy.kord at googlemail.com (Ted Kord)
Date: Tue, 9 Sep 2008 16:28:21 -0700 (PDT)
Subject: [SciPy-user] Pytrilinos on Windows
In-Reply-To: <48C6D8AD.4080608@sun.ac.za>
References: <34d365e70809091254x7ec84543y13b032aae141cd6b@mail.gmail.com> 
	<48C6D8AD.4080608@sun.ac.za>
Message-ID: <1bbf2130-7eae-45f1-a504-c54c5462db7d@m45g2000hsb.googlegroups.com>

I'll have a look. Thanks.

Ted


From oliphant at enthought.com  Wed Sep 10 00:38:24 2008
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Tue, 09 Sep 2008 23:38:24 -0500
Subject: [SciPy-user] NumPy arrays that use memory allocated from other
	libraries or tools
Message-ID: <48C74F40.3090103@enthought.com>


I wanted to point anybody interested to a blog post that describes a 
useful pattern for having a NumPy array that points to the memory 
created by a different memory manager than the standard one used by 
NumPy.   The pattern  shows how to create a NumPy array that points to 
previously allocated memory and then shows how to construct an object 
that allows the correct deallocator to be called when the NumPy array is 
freed.

This may be useful if you are wrapping code that has it's own memory 
management scheme.   Comments and feedback is welcome. 

The post is

http://blog.enthought.com/?p=62


Best regards,

-Travis Oliphant



From simon.palmer at gmail.com  Wed Sep 10 20:41:59 2008
From: simon.palmer at gmail.com (SimonPalmer)
Date: Wed, 10 Sep 2008 17:41:59 -0700 (PDT)
Subject: [SciPy-user] examples of using norm_gen
Message-ID: <510dfcba-748d-4541-91e5-a6fb340bc046@p31g2000prf.googlegroups.com>

can anyone point me in the direction of examples of how to use norm/
norm_gen from the scipy.stats.distributions module?
thanks
Simon


From robert.kern at gmail.com  Thu Sep 11 00:05:30 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 10 Sep 2008 23:05:30 -0500
Subject: [SciPy-user] examples of using norm_gen
In-Reply-To: <510dfcba-748d-4541-91e5-a6fb340bc046@p31g2000prf.googlegroups.com>
References: <510dfcba-748d-4541-91e5-a6fb340bc046@p31g2000prf.googlegroups.com>
Message-ID: <3d375d730809102105v167195ceuf8c0cf9e725c5059@mail.gmail.com>

On Wed, Sep 10, 2008 at 19:41, SimonPalmer <simon.palmer at gmail.com> wrote:
> can anyone point me in the direction of examples of how to use norm/
> norm_gen from the scipy.stats.distributions module?

The docstring gives a good overview of its capabilities. Is there
something specific you found confusing or inadequate about it?

-- 
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 simon.palmer at gmail.com  Thu Sep 11 00:17:27 2008
From: simon.palmer at gmail.com (SimonPalmer)
Date: Wed, 10 Sep 2008 21:17:27 -0700 (PDT)
Subject: [SciPy-user] examples of using norm_gen
In-Reply-To: <3d375d730809102105v167195ceuf8c0cf9e725c5059@mail.gmail.com>
References: <510dfcba-748d-4541-91e5-a6fb340bc046@p31g2000prf.googlegroups.com>
	<3d375d730809102105v167195ceuf8c0cf9e725c5059@mail.gmail.com>
Message-ID: <aab1ed58-8c24-4c6f-b394-0c443a080fa6@t1g2000pra.googlegroups.com>

Inadequate, certainly not.  Confusing yes.  However, I think my
problem is really a lack of experience with python.  I have read the
docstrings,  and looking at what the module does it seems to suit my
purposes perfectly, I'm slightly embarrassed to say that my lack of
python skills means I am baffled about how I would actually use it,
hence the request to see some sample code.

I'm missing something obvious, I can't blame the module.

On Sep 11, 5:05?am, "Robert Kern" <robert.k... at gmail.com> wrote:
> On Wed, Sep 10, 2008 at 19:41, SimonPalmer <simon.pal... at gmail.com> wrote:
> > can anyone point me in the direction of examples of how to use norm/
> > norm_gen from the scipy.stats.distributions module?
>
> The docstring gives a good overview of its capabilities. Is there
> something specific you found confusing or inadequate about it?
>
> --
> Robert Kern
>
> "I have come to believe that the whole world is an enigma, a harmless
> enigma that is made terrible by our own mad attempt to interpret it as
> though it had an underlying truth."
> ?-- Umberto Eco
> _______________________________________________
> SciPy-user mailing list
> SciPy-u... at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user


From tonyyu at MIT.EDU  Thu Sep 11 11:03:06 2008
From: tonyyu at MIT.EDU (Tony S Yu)
Date: Thu, 11 Sep 2008 11:03:06 -0400
Subject: [SciPy-user] Unexpected typecasting when adding sparse.lil_matrix
Message-ID: <DA650C36-9353-46C7-9B15-6D0A3F569B42@mit.edu>

This may be expected behavior, but I found it surprising. Addition (or  
any other simple operation) of two lil sparse matrices returns a csc  
sparse matrix. The scipy website suggests that csc and csr matrices  
are more efficient than lil matrices for multiplication and inversion  
(and I guess for addition too), but this typecasting is still a little  
surprising. I was just curious if this is intentional.

Thanks,
-Tony

#~~~~~~~~~

In [1]: import scipy.sparse as sparse

In [2]: A = sparse.lil_eye([3, 3])

In [3]: A + A

Out[3]:
<3x3 sparse matrix of type '<type 'numpy.float64'>'
	with 3 stored elements (space for 3)
	in Compressed Sparse Column format>



From wnbell at gmail.com  Thu Sep 11 11:49:02 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Thu, 11 Sep 2008 11:49:02 -0400
Subject: [SciPy-user] Unexpected typecasting when adding
	sparse.lil_matrix
In-Reply-To: <DA650C36-9353-46C7-9B15-6D0A3F569B42@mit.edu>
References: <DA650C36-9353-46C7-9B15-6D0A3F569B42@mit.edu>
Message-ID: <d05265cb0809110849g275a2ad7y204433077871cfb4@mail.gmail.com>

On Thu, Sep 11, 2008 at 11:03 AM, Tony S Yu <tonyyu at mit.edu> wrote:
> This may be expected behavior, but I found it surprising. Addition (or
> any other simple operation) of two lil sparse matrices returns a csc
> sparse matrix. The scipy website suggests that csc and csr matrices
> are more efficient than lil matrices for multiplication and inversion
> (and I guess for addition too), but this typecasting is still a little
> surprising. I was just curious if this is intentional.
>
> Thanks,
> -Tony
>
> #~~~~~~~~~
>
> In [1]: import scipy.sparse as sparse
>
> In [2]: A = sparse.lil_eye([3, 3])
>
> In [3]: A + A
>
> Out[3]:
> <3x3 sparse matrix of type '<type 'numpy.float64'>'
>        with 3 stored elements (space for 3)
>        in Compressed Sparse Column format>
>

That is the intended result (currently).

The issue here is that only some sparse formats define arithmetic
operations, so those that don't are converted to a type that does.
Even if lil_matrix did define addition itself, it would almost
certainly be slower than conversion to a "native" format (i.e. one
with a lower-level C++ implementation).  Furthermore, guaranteeing
that other operations, such as A.transpose(), return the same type as
A would make csr_matrix.transpose() an O(N) method rather than an O(1)
method.

The current implementation is nice because inexperienced users will
often get better performance in subsequent operations.  OTOH, as your
case illustrates, it can lead to surprises.

You can always do (A + B).tolil() if you want the result in a
particular format though.  I'm not sure if your version of SciPy
supports lil_matrix(A + B), but that works now also.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From bryan at ideotrope.org  Thu Sep 11 17:33:58 2008
From: bryan at ideotrope.org (Bryan Keith)
Date: Thu, 11 Sep 2008 15:33:58 -0600 (MDT)
Subject: [SciPy-user] scipy.test fails: clapack module is empty (Bryan
 Keith)
In-Reply-To: <48C6AC36.6030808@meduni-graz.at>
References: <mailman.2013.1220978900.3376.scipy-user@scipy.org>
	<48C6AC36.6030808@meduni-graz.at>
Message-ID: <4528.64.78.232.178.1221168838.squirrel@ideotrope.org>

Bernardo,

I have no idea what to do about this error.  I was hoping someone on this
list might be able to help...

Bryan

> Hi Bryan Keith,
>>    8. scipy.test fails: clapack module is empty (Bryan Keith)
>>
> I've got the same error with my installation (opensuse11 via yast ---
> science repository), please have a look at the message below. I also
> have a "Illegal Instruction" error in the end of the test, which is
> worse than your problem. I would like to get rid off this. Any
> suggestions?
>
> Thanks!
>
> --------------------------
> In [1]: import scipy
>
> In [2]: scipy.test()
> Failed importing scipy.linsolve.umfpack: 'module' object has no
> attribute 'umfpack'
> Found 9/9 tests for scipy.cluster.tests.test_vq
> Found 20/20 tests for scipy.fftpack.tests.test_pseudo_diffs
> Found 4/4 tests for scipy.fftpack.tests.test_helper
> Found 18/18 tests for scipy.fftpack.tests.test_basic
> Found 1/1 tests for scipy.integrate.tests.test_integrate
> Found 3/3 tests for scipy.integrate.tests.test_quadrature
> Found 10/10 tests for scipy.integrate.tests.test_quadpack
> Found 6/6 tests for scipy.tests.test_fitpack
> Found 6/6 tests for scipy.tests.test_interpolate
> Found 28/28 tests for scipy.io.tests.test_mio
> Found 4/4 tests for scipy.io.tests.test_recaster
> Found 5/5 tests for scipy.io.tests.test_npfile
> Found 13/13 tests for scipy.io.tests.test_mmio
> Found 4/4 tests for scipy.io.tests.test_array_import
> Found 16/16 tests for scipy.lib.blas.tests.test_blas
> Found 128/128 tests for scipy.lib.blas.tests.test_fblas
>
> ****************************************************************
> WARNING: clapack module is empty
> -----------
> See scipy/INSTALL.txt for troubleshooting.
> Notes:
> * If atlas library is not found by numpy/distutils/system_info.py,
> then scipy uses flapack instead of clapack.
> ****************************************************************
>
> Found 42/42 tests for scipy.lib.lapack.tests.test_lapack
> Found 4/4 tests for scipy.linalg.tests.test_lapack
> <module 'scipy.linalg.fblas' from
> '/usr/lib64/python2.5/site-packages/scipy/linalg/fblas.so'>
> Found 16/16 tests for scipy.linalg.tests.test_blas
> Found 6/6 tests for scipy.linalg.tests.test_iterative
> Found 41/41 tests for scipy.linalg.tests.test_basic
> Found 128/128 tests for scipy.linalg.tests.test_fblas
> Found 7/7 tests for scipy.linalg.tests.test_matfuncs
> Found 72/72 tests for scipy.linalg.tests.test_decomp
> Failed importing
> /usr/lib64/python2.5/site-packages/scipy/linsolve/umfpack/tests/test_umfpack.py:
> 'module' object has no attribute 'umfpack'
> Found 2/2 tests for scipy.maxentropy.tests.test_maxentropy
> Found 3/3 tests for scipy.misc.tests.test_pilutil
> Found 399/399 tests for scipy.ndimage.tests.test_ndimage
> Found 5/5 tests for scipy.odr.tests.test_odr
> Found 4/4 tests for scipy.optimize.tests.test_zeros
> Found 1/1 tests for scipy.optimize.tests.test_cobyla
> Found 8/8 tests for scipy.optimize.tests.test_optimize
> Found 10/10 tests for scipy.optimize.tests.test_nonlin
> Found 5/5 tests for scipy.signal.tests.test_signaltools
> Found 4/4 tests for scipy.signal.tests.test_wavelets
> Found 152/152 tests for scipy.sparse.tests.test_sparse
> Found 3/3 tests for scipy.special.tests.test_spfun_stats
> Found 342/342 tests for scipy.special.tests.test_basic
> Found 10/10 tests for scipy.stats.tests.test_morestats
> Found 107/107 tests for scipy.stats.tests.test_stats
> Found 73/73 tests for scipy.stats.tests.test_distributions
> Found 0/0 tests for scipy.weave.tests.test_c_spec
> Found 1/1 tests for scipy.weave.tests.test_ast_tools
> Found 9/9 tests for scipy.weave.tests.test_build_tools
> Found 2/2 tests for scipy.weave.tests.test_blitz_tools
> building extensions here: /home/rocha/.python25_compiled/m11
> Found 1/1 tests for scipy.weave.tests.test_ext_tools
> Found 0/0 tests for scipy.weave.tests.test_inline_tools
> Found 26/26 tests for scipy.weave.tests.test_catalog
> Found 0/0 tests for scipy.weave.tests.test_scxx_sequence
> Found 74/74 tests for scipy.weave.tests.test_size_check
> Failed importing
> /usr/lib64/python2.5/site-packages/scipy/weave/tests/old_test_wx_spec.py:
> Could not locate wxPython base directory.
> Found 0/0 tests for scipy.weave.tests.test_scxx_dict
> Found 0/0 tests for scipy.weave.tests.test_scxx_object
> Found 3/3 tests for scipy.weave.tests.test_standard_array_spec
> Found 16/16 tests for scipy.weave.tests.test_slice_handler
> Failed importing
> /usr/lib64/python2.5/site-packages/scipy/weave/tests/test_wx_spec.py:
> Could not locate wxPython base directory.
> ../usr/lib64/python2.5/site-packages/scipy/cluster/vq.py:477:
> UserWarning: One of the clusters is empty. Re-run kmean with a different
> initialization.
> warnings.warn("One of the clusters is empty. "
> exception raised as expected: One of the clusters is empty. Re-run kmean
> with a different initialization.
> ..........Illegal instruction
>
> Bernardo M. Rocha
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>




From mnandris at blueyonder.co.uk  Fri Sep 12 07:02:02 2008
From: mnandris at blueyonder.co.uk (Michael)
Date: Fri, 12 Sep 2008 12:02:02 +0100
Subject: [SciPy-user] Why doesn't norm_gen have a 'dist' attribute?
In-Reply-To: <mailman.13.1221152403.31955.scipy-user@scipy.org>
References: <mailman.13.1221152403.31955.scipy-user@scipy.org>
Message-ID: <1221217322.6305.6.camel@mik>


> Message: 2
> Date: Wed, 10 Sep 2008 23:05:30 -0500
> From: "Robert Kern" <robert.kern at gmail.com>
> Subject: Re: [SciPy-user] examples of using norm_gen
> To: "SciPy Users List" <scipy-user at scipy.org>
> Message-ID:
> 	<3d375d730809102105v167195ceuf8c0cf9e725c5059 at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
> 
> On Wed, Sep 10, 2008 at 19:41, SimonPalmer <simon.palmer at gmail.com> wrote:
> > can anyone point me in the direction of examples of how to use norm/
> > norm_gen from the scipy.stats.distributions module?
> 
> The docstring gives a good overview of its capabilities. Is there
> something specific you found confusing or inadequate about it?

norm_gen doesn't appear to _generate_ anything, not directly anyway

a=d.norm_gen(name='norm',longname='a normal') 
a is <class 'scipy.stats.distributions.norm_gen'>
b=d.norm(x,size=n) 				
b is <class 'scipy.stats.distributions.rv_frozen'>

both a and b produce normal distributions that don't 'look' normal - see
attached, though this is a bad test since the beta distribution looks
totally mangled but is correc

pdf1=a.pdf(x,size=n)
pdf2=b.dist.pdf(x,size=n) 

Leaving norm_gen aside for a moment, c.f. the lack of the 'dist'
attribute in 'a', one might also ask: what the dist attribute does?
Where did it spring from? there's too much padding in the scipy
sandwich, so to speak;

e.g. the sheer number of ways of creating scipy distributions is
slightly baffling; afaik norm_gen is a base class that is not to be used
directly... but that's just a guess.

That said, there is an 85-100 fold speed-up over using a np.random
distribution + np.histogram

Is there a guide somewhere explaining how scientific python api's are
packaged/structured? import idioms?


> -- 
> 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

-------------- next part --------------
A non-text attachment was scrubbed...
Name: norm.py
Type: text/x-python
Size: 489 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080912/473bda9a/attachment.py>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080912/473bda9a/attachment.sig>

From arserlom at gmail.com  Fri Sep 12 08:09:32 2008
From: arserlom at gmail.com (Armando Serrano Lombillo)
Date: Fri, 12 Sep 2008 14:09:32 +0200
Subject: [SciPy-user] Why doesn't norm_gen have a 'dist' attribute?
In-Reply-To: <1221217322.6305.6.camel@mik>
References: <mailman.13.1221152403.31955.scipy-user@scipy.org>
	<1221217322.6305.6.camel@mik>
Message-ID: <d7ab268c0809120509o3494cef0o6d6d47081601c710@mail.gmail.com>

Hello Michael. norm_gen is not meant to be used, it is just used inside
scipy's code to generate the norm class, which you should use.

For example:

from scipy.stats import norm

norm.rvs(size=n)
norm.pdf(x)

or

d = norm(loc=mean)
d.pdf(x)

or

norm.pdf(x, loc=mean)

or whatever you need to do.

Hope it helps,
Armando.

On Fri, Sep 12, 2008 at 1:02 PM, Michael <mnandris at blueyonder.co.uk> wrote:

>
> > Message: 2
> > Date: Wed, 10 Sep 2008 23:05:30 -0500
> > From: "Robert Kern" <robert.kern at gmail.com>
> > Subject: Re: [SciPy-user] examples of using norm_gen
> > To: "SciPy Users List" <scipy-user at scipy.org>
> > Message-ID:
> >       <3d375d730809102105v167195ceuf8c0cf9e725c5059 at mail.gmail.com>
> > Content-Type: text/plain; charset=UTF-8
> >
> > On Wed, Sep 10, 2008 at 19:41, SimonPalmer <simon.palmer at gmail.com>
> wrote:
> > > can anyone point me in the direction of examples of how to use norm/
> > > norm_gen from the scipy.stats.distributions module?
> >
> > The docstring gives a good overview of its capabilities. Is there
> > something specific you found confusing or inadequate about it?
>
> norm_gen doesn't appear to _generate_ anything, not directly anyway
>
> a=d.norm_gen(name='norm',longname='a normal')
> a is <class 'scipy.stats.distributions.norm_gen'>
> b=d.norm(x,size=n)
> b is <class 'scipy.stats.distributions.rv_frozen'>
>
> both a and b produce normal distributions that don't 'look' normal - see
> attached, though this is a bad test since the beta distribution looks
> totally mangled but is correc
>
> pdf1=a.pdf(x,size=n)
> pdf2=b.dist.pdf(x,size=n)
>
> Leaving norm_gen aside for a moment, c.f. the lack of the 'dist'
> attribute in 'a', one might also ask: what the dist attribute does?
> Where did it spring from? there's too much padding in the scipy
> sandwich, so to speak;
>
> e.g. the sheer number of ways of creating scipy distributions is
> slightly baffling; afaik norm_gen is a base class that is not to be used
> directly... but that's just a guess.
>
> That said, there is an 85-100 fold speed-up over using a np.random
> distribution + np.histogram
>
> Is there a guide somewhere explaining how scientific python api's are
> packaged/structured? import idioms?
>
>
> > --
> > Robert Kern
> >
> > "I have come to believe that the whole world is an enigma, a harmless
> > enigma that is made terrible by our own mad attempt to interpret it as
> > though it had an underlying truth."
> >  -- Umberto Eco
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080912/e81cca0b/attachment.html>

From josef.pktd at gmail.com  Fri Sep 12 11:17:07 2008
From: josef.pktd at gmail.com (joep)
Date: Fri, 12 Sep 2008 08:17:07 -0700 (PDT)
Subject: [SciPy-user] **kwds in frozen distribution class
Message-ID: <d28d14ca-48d9-4d3e-b053-1c718fd23bda@m36g2000hse.googlegroups.com>

I was trying to work on an example for the frozen distribution class,
in response to the previous message by Michael.

The change in method signature between frozen and not frozen class got
me
confused:
I looked at the information in
help(stats.norm.stats)

and did not realize that the signature in
help(stats.norm(loc = 10, scale = 10).stats) or help(stats.norm(loc =
10, scale = 10))

is different.

Given the description it does what it says, but I think it would be
easy to
make the keyword arguments consistent between frozen and not frozen
distribution


The frozen distribution class does not take any additional keywords
-------------------------------------------------------------------
e.g.

>>> from scipy import stats
>>> stats.norm.stats(loc = 10, scale = 10, moments='v')
array(100.0)
>>> stats.norm(loc = 10, scale = 10).stats(moments='v')
Traceback (most recent call last):
  File "<pyshell#91>", line 1, in ?
    stats.norm(loc = 10, scale = 10).stats(moments='v')
TypeError: stats() got an unexpected keyword argument 'moments'
>>> stats.norm(loc = 10, scale = 10, moments='v').stats()
array(100.0)


line numbers from current trunk
(http://projects.scipy.org/scipy/scipy/browser/trunk/scipy/stats/
distributions.py)

101 	# Frozen RV class
102 	class rv_frozen(object):

121 	    def stats(self):
122 	        return self.dist.stats(*self.args,**self.kwds)

I think this change should work to accept additional keywords:

121 	    def stats(self,**kwds):
                kwds.update(self.kwds)
122 	        return self.dist.stats(*self.args,**kwds)

Josef


From josef.pktd at gmail.com  Fri Sep 12 15:42:39 2008
From: josef.pktd at gmail.com (joep)
Date: Fri, 12 Sep 2008 12:42:39 -0700 (PDT)
Subject: [SciPy-user] **kwds in frozen distribution class
In-Reply-To: <d28d14ca-48d9-4d3e-b053-1c718fd23bda@m36g2000hse.googlegroups.com>
References: <d28d14ca-48d9-4d3e-b053-1c718fd23bda@m36g2000hse.googlegroups.com>
Message-ID: <10729024-7252-47f7-83e8-7dc4a28ff4e1@w7g2000hsa.googlegroups.com>

problems with **kwds in moment method in distributions
======================================================

Summary:
--------

* ``moment`` in frozen continuous distribution with loc and scale
keyword arguments raises exception
* ``moment`` in frozen discrete distribution accepts the presence of
loc and scale, but ignores them
* ``stats`` method works correctly, but only for first and second
moment

>>> scipy.version.version
'0.6.0'
>>> numpy.version.version
'1.1.0'

example continuous distribution
-------------------------------


>>> stats.gamma(4).stats()
(array(4.0), array(4.0))
>>> stats.gamma(4).moment(1)
4
>>> stats.gamma(4).moment(2)
4

moment in frozen distribution with loc and scale keyword arguments
raises exception

>>> stats.gamma(4,loc = 10, scale = 10).stats()
(array(50.0), array(400.0))
>>> stats.gamma(4,loc = 10, scale = 10).moment(1)
Traceback (most recent call last):
  File "<pyshell#166>", line 1, in ?
    stats.gamma(4,loc = 10, scale = 10).moment(1)
  File "C:\Programs\Python24\lib\site-packages\scipy\stats
\distributions.py", line 124, in moment
    return self.dist.moment(n,*self.args,**self.kwds)
TypeError: moment() got an unexpected keyword argument 'loc'


actually moment does not allow for loc and scale in the not-frozen
distribution
either:

>>> stats.gamma.moment(2,4)
4
>>> stats.gamma.stats(4,loc = 10, scale = 10)
(array(50.0), array(400.0))

>>> stats.gamma.moment(2,4,loc = 10, scale = 10)
Traceback (most recent call last):
  File "<pyshell#175>", line 1, in ?
    stats.gamma.moment(2,4,loc = 10, scale = 10)
TypeError: moment() got an unexpected keyword argument 'loc'


check:stats agrees with random sample:

>>> rvs=stats.gamma.rvs(4,loc = 10, scale = 10,size=10000)
>>> rvs.mean()
50.01717741991262
>>> rvs.var()
400.12936828905185


example: discrete distribution
------------------------------

stats works correctly with or without loc,scale parameters:

>>> stats.poisson.stats(4,loc = 10, scale = 10)
(array(14.0), array(4.0))
>>> stats.poisson(4,loc = 10, scale = 10).stats()
(array(14.0), array(4.0))
>>> rvs=stats.poisson.rvs(4,loc = 10, scale = 10, size=10000)
>>> rvs.mean()
13.9985
>>> rvs.var()
3.9602977499998575

>>> stats.poisson.stats(4)
(array(4.0), array(4.0))
>>> stats.poisson(4).stats()
(array(4.0), array(4.0))

moment in frozen distribution accepts the presence of loc and scale,
but ignores them

the following are not for the log,scale transformed random variable
>>> stats.poisson(4,loc = 10, scale = 10).moment(1)  #wrong result
4
>>> stats.poisson(4,loc = 10, scale = 10).moment(2)  #wrong result
4
>>> stats.poisson(4,loc = 10, scale = 10).moment(3)  #wrong result
4.0
>>> stats.poisson(4,loc = 10, scale = 10).moment(4)  #wrong result
52.0
>>> stats.poisson.moment(1,4,loc = 10, scale = 10)  #wrong result
4

instead the results are for the untransformed variable

>>> stats.poisson(4).moment(1)
4
>>> stats.poisson(4).moment(2)
4
>>> stats.poisson(4).moment(3)
4.0
>>> stats.poisson(4).moment(4)
52.0

just to check result
>>> rvs0=stats.poisson.rvs(4,size=10000000)
>>> ((rvs0-4)**4).mean()
51.991537899999997

adding loc, scale to moment of already frozen distribution raises
exception,
which at least is not misleading

>>> stats.poisson(4).moment(1,loc = 10, scale = 10)
Traceback (most recent call last):
  File "<pyshell#227>", line 1, in ?
    stats.poisson(4).moment(1,loc = 10, scale = 10)
TypeError: moment() got an unexpected keyword argument 'loc'


Josef


From ryanlists at gmail.com  Fri Sep 12 17:18:21 2008
From: ryanlists at gmail.com (Ryan Krauss)
Date: Fri, 12 Sep 2008 16:18:21 -0500
Subject: [SciPy-user] butterworth filter
In-Reply-To: <48C67C16.9030701@free.fr>
References: <48C67C16.9030701@free.fr>
Message-ID: <c5b438120809121418u3ca83865o521271be3c67f70c@mail.gmail.com>

So, attached is the plot I get from scipy.  It is a high pass filter.  It
seems reasonable.  What does the curve look like from Matlab.

Ryan

On Tue, Sep 9, 2008 at 8:37 AM, cyril giraudon <cyril.giraudon at free.fr>wrote:

> Hi,
>
> I use scipy 0.6.0 and i try to reproduce the plot of the matlab butter
> function web documentation (first response for a google request "matlab
> butter example").
>
> The matlab code is :
>
> [z,p,k] = butter(9,300/500,'high');
> [sos,g] = zp2sos(z,p,k);             % Convert to SOS form
> Hd = dfilt.df2tsos(sos,g);   % Create a dfilt object
> h = fvtool(Hd);              % Plot magnitude response
> set(h,'Analysis','freq')             % Display frequency response
>
>
> In scipy, I write :
>
> from scipy.signal import butter, freqz
>
> from pylab import show, grid, log, plot
>
> b, a = butter(9, 300./500., 'high')
>
> fi = freqz(b, a)
>
> plot(fi[0], 20*log(abs(fi[1])))
>
> grid()
>
> show()
>
>
>
> Why are the two filters not the same ?
>
>
> Thanks a lot,
>
> Cyril.
>
>
> _______________________________________________
> 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/20080912/42c1edb7/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: butter_scipy.png
Type: image/png
Size: 33022 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080912/42c1edb7/attachment.png>

From cournape at gmail.com  Sat Sep 13 01:35:14 2008
From: cournape at gmail.com (David Cournapeau)
Date: Sat, 13 Sep 2008 14:35:14 +0900
Subject: [SciPy-user] scipy.test fails: clapack module is empty (Bryan
	Keith)
In-Reply-To: <4528.64.78.232.178.1221168838.squirrel@ideotrope.org>
References: <mailman.2013.1220978900.3376.scipy-user@scipy.org>
	<48C6AC36.6030808@meduni-graz.at>
	<4528.64.78.232.178.1221168838.squirrel@ideotrope.org>
Message-ID: <5b8d13220809122235w1a5ecc7do57552adbdb9f5f91@mail.gmail.com>

On Fri, Sep 12, 2008 at 6:33 AM, Bryan Keith <bryan at ideotrope.org> wrote:
> Bernardo,
>
> I have no idea what to do about this error.  I was hoping someone on this
> list might be able to help...

It is likely to be an error with blas/lapack/atlas as built in the
repository. Some other people use suse I believe, maybe they will be
able to help.

cheers,

David


From f.braennstroem at gmx.de  Sat Sep 13 09:39:23 2008
From: f.braennstroem at gmx.de (Fabian Braennstroem)
Date: Sat, 13 Sep 2008 15:39:23 +0200
Subject: [SciPy-user] scale arrays with interpolation
Message-ID: <gagfo8$6b3$1@ger.gmane.org>


Hi,


in fluid dynamics one often plots velocity distribution scaled by some
reference to take a look at self-similarity. E.g. for free jets one can
extract the velocity across the plot, which gives
two different arrays; an x_array and an y_array, where the x_array corresponds
to the position and the y_array to the velocity at this position. So 
one would get two arrays of the same size:

x_array=[ 1,2,3,4,5,6,7,8,9,10]
y_array=[ 0,1,1,2,5,2,1,1,1,1]

For the visualization of self-similar behavior one scales the y_array with the
local maximum (in this case 5) and the x_array with the position, where the
velocity is the half of the maximum. This means one has to calculate the half
velocity (5/2=2.5) and has to interpolate the position of this value to get the
position (in this case it is somewhere between 4 and 5). This half-width
position would be used to scale the x_array. At the end one gets:

x_array_scaled= [1/half_width_position, 2/half_with_position,...,10/half_width_position]
y_array_scaled= [0/local_max,1/local_max,...,1/local_max]

Does anyone have a suggestion how to do this? The major problem is how to get
the 'half_width_position'. Would be nice, if anyone has an idea!


Greetings!
 Fabian



From alex.liberzon at gmail.com  Sat Sep 13 16:35:21 2008
From: alex.liberzon at gmail.com (Alex Liberzon)
Date: Sat, 13 Sep 2008 22:35:21 +0200
Subject: [SciPy-user] scale arrays with interpolation
Message-ID: <48CC2409.1010501@gmail.com>

Hi Fabian

Fortunately, I'm from the same field so I think I understand what you're 
talking about :-) If I translate your wish is::
a) find the location and the value of the maximum of the y_array => 
local_max
b) half_width is the x_array value at the position of the maximum of y_array

If this is true, then you'd better use numpy arrays and not lists:
# starting from your lists:
 x_array = asarray(x_array).astype('f') # array of floats
 y_array = asarray(y_array).astype('f')

half_width = x_array[argmax(y_array)]

scaled_x_array = x_array/half_width
scaled_y_array = y_array/max(y_array)

plot(scaled_x_array,scaled_y_array,'o')

Hope it helps,
Alex
 



From f.braennstroem at gmx.de  Sun Sep 14 05:29:04 2008
From: f.braennstroem at gmx.de (Fabian Braennstroem)
Date: Sun, 14 Sep 2008 11:29:04 +0200
Subject: [SciPy-user] scale arrays with interpolation
References: <48CC2409.1010501@gmail.com>
Message-ID: <gailer$v5f$1@ger.gmane.org>

Hi Alex,

* Alex Liberzon <alex.liberzon at gmail.com> wrote:
> Hi Fabian
>
> Fortunately, I'm from the same field so I think I understand what you're 
> talking about :-) 

I hope it was not too confusing...

> If I translate your wish is::
> a) find the location and the value of the maximum of the y_array => 
> local_max
> b) half_width is the x_array value at the position of the maximum of y_array
>
> If this is true, then you'd better use numpy arrays and not lists:
> # starting from your lists:
>  x_array = asarray(x_array).astype('f') # array of floats
>  y_array = asarray(y_array).astype('f')
>
> half_width = x_array[argmax(y_array)]
>
> scaled_x_array = x_array/half_width
> scaled_y_array = y_array/max(y_array)
>
> plot(scaled_x_array,scaled_y_array,'o')

Thanks! This is a good way with a lot of elements, but I
think, if one does this with just a few elements in the
numpy array, one has to do some kind of interpolation, to
find the 'exact'
half_width!?

Regards!
 Fabian



From robince at gmail.com  Sun Sep 14 05:38:32 2008
From: robince at gmail.com (Robin)
Date: Sun, 14 Sep 2008 10:38:32 +0100
Subject: [SciPy-user] scale arrays with interpolation
In-Reply-To: <gailer$v5f$1@ger.gmane.org>
References: <48CC2409.1010501@gmail.com> <gailer$v5f$1@ger.gmane.org>
Message-ID: <b14b627d0809140238g2568f42bpd214f5f132527df6@mail.gmail.com>

On Sun, Sep 14, 2008 at 10:29 AM, Fabian Braennstroem
<f.braennstroem at gmx.de> wrote:

> Thanks! This is a good way with a lot of elements, but I
> think, if one does this with just a few elements in the
> numpy array, one has to do some kind of interpolation, to
> find the 'exact'
> half_width!?

You can find interpolation functions at numpy.interp and in the
scipy.interpolate module.
 Probably scipy.interpolate.interp1d is what you need, but I'm not
sure of the difference between that and numpy.interp.

Robin


From simon.palmer at gmail.com  Sun Sep 14 06:50:45 2008
From: simon.palmer at gmail.com (SimonPalmer)
Date: Sun, 14 Sep 2008 03:50:45 -0700 (PDT)
Subject: [SciPy-user] step size using optimize.fmin
Message-ID: <bf9a6e1f-db3d-4edf-8ac4-e72dbcdf21ac@m45g2000hsb.googlegroups.com>

Hi, I have a weird and lumpy N-D function that I am trying to minimize
using optimize.fmin.  The problem I am having is that the step size
that fmin uses is fixed and not quite big enough to move the function
value sufficiently far to examine a different "state", so the function
terminates (cleanly) presuming it is on a flat surface.

I have been looking through the code and I can't see a way of
adjusting the step size, which seem to be magic numbers in the code:

    nonzdelt = 0.05
    zdelt = 0.00025

I guess what I am looking for is one of:
a) a way to set these to be a slightly larger value
b) a callback to be able to adjust them if the convergence criteria
are met but nothing has changed (i.e. a flat surface)
c) a different minimmisation algorithm which has momentum

I am also contemplating

d) change my function so that it is somewhat more continuous in nature
e) hack a local copy of fmin

but I would rather not do either unless I have to.

Anyone have any recommendations?
Am I reading fmin right?

tia
Simon


From f.braennstroem at gmx.de  Sun Sep 14 06:55:21 2008
From: f.braennstroem at gmx.de (Fabian Braennstroem)
Date: Sun, 14 Sep 2008 12:55:21 +0200
Subject: [SciPy-user] scale arrays with interpolation
References: <48CC2409.1010501@gmail.com> <gailer$v5f$1@ger.gmane.org>
	<b14b627d0809140238g2568f42bpd214f5f132527df6@mail.gmail.com>
Message-ID: <gaiqgl$ce0$1@ger.gmane.org>

Hi Robin,

* Robin <robince at gmail.com> wrote:
> On Sun, Sep 14, 2008 at 10:29 AM, Fabian Braennstroem
> <f.braennstroem at gmx.de> wrote:
>
>> Thanks! This is a good way with a lot of elements, but I
>> think, if one does this with just a few elements in the
>> numpy array, one has to do some kind of interpolation, to
>> find the 'exact'
>> half_width!?
>
> You can find interpolation functions at numpy.interp and in the
> scipy.interpolate module.
>  Probably scipy.interpolate.interp1d is what you need, but I'm not
> sure of the difference between that and numpy.interp.

Thanks, this should work!

Greetings!
 Fabian



From pav at iki.fi  Sun Sep 14 07:34:18 2008
From: pav at iki.fi (Pauli Virtanen)
Date: Sun, 14 Sep 2008 11:34:18 +0000 (UTC)
Subject: [SciPy-user] step size using optimize.fmin
References: <bf9a6e1f-db3d-4edf-8ac4-e72dbcdf21ac@m45g2000hsb.googlegroups.com>
Message-ID: <gaisrq$i1j$1@ger.gmane.org>

Sun, 14 Sep 2008 03:50:45 -0700, SimonPalmer wrote:

> Hi, I have a weird and lumpy N-D function that I am trying to minimize
> using optimize.fmin.  The problem I am having is that the step size that
> fmin uses is fixed and not quite big enough to move the function value
> sufficiently far to examine a different "state", so the function
> terminates (cleanly) presuming it is on a flat surface.

Silly thought: maybe you could scale your function so that its 
characteristic length scale becomes 1?

x_scale = 1000
f_scale = 1

def scaled_func(x):
    return func(x * scale) * f_scale

xopt = sp.optimize.fmin(scaled_func, scale*x0)
xopt /= scale


-- 
Pauli Virtanen



From josef.pktd at gmail.com  Sun Sep 14 09:52:27 2008
From: josef.pktd at gmail.com (joep)
Date: Sun, 14 Sep 2008 06:52:27 -0700 (PDT)
Subject: [SciPy-user] ipython install: egg not python 2.4 compatible
Message-ID: <a700e94c-2972-498a-8de8-5dd7b9a0cf7a@z66g2000hsc.googlegroups.com>

Sorry for wrong group: I'm not subscribed to ipython-user:

ipython-0.9-py2.4.egg uses features from python 2.5:
error messages with easy_install::

C:\Programs\Python24\Scripts\easy_install-script.py -U C:\Josef\work-
oth\sort\pypi\ipython-0.9-py2.4.egg
Processing ipython-0.9-py2.4.egg
creating c:\programs\python24\lib\site-packages\ipython-0.9-py2.4.egg
Extracting ipython-0.9-py2.4.egg to c:\programs\python24\lib\site-
packages
  File "c:\programs\python24\lib\site-packages\ipython-0.9-py2.4.egg
\IPython\con
fig\config.py", line 49
    with raw(self):
           ^
SyntaxError: invalid syntax
  File "c:\programs\python24\lib\site-packages\ipython-0.9-py2.4.egg
\IPython\fro
ntend\linefrontendbase.py", line 192
    finally:
          ^
SyntaxError: invalid syntax
  File "c:\programs\python24\lib\site-packages\ipython-0.9-py2.4.egg
\IPython\fro
ntend\prefilterfrontend.py", line 207
    finally:
          ^
SyntaxError: invalid syntax
SyntaxError: ('future feature with_statement is not defined',)
  File "c:\programs\python24\lib\site-packages\ipython-0.9-py2.4.egg
\IPython\ker
nel\tests\test_contexts.py", line 28
    with parallel as pr:
                ^
SyntaxError: invalid syntax
  File "c:\programs\python24\lib\site-packages\ipython-0.9-py2.4.egg
\share\doc\i
python\examples\kernel\nwmerge.py", line 48
    toadd = (key(item), i, item, itr) if key else (item, i, itr)
                                       ^
SyntaxError: invalid syntax
Adding ipython 0.9 to easy-install.pth file
Installing iptest-script.py script to C:\Programs\Python24\Scripts
Installing iptest.exe script to C:\Programs\Python24\Scripts
Installing ipythonx-script.py script to C:\Programs\Python24\Scripts
Installing ipythonx.exe script to C:\Programs\Python24\Scripts
Installing ipcluster-script.py script to C:\Programs\Python24\Scripts
Installing ipcluster.exe script to C:\Programs\Python24\Scripts
Installing ipython-script.py script to C:\Programs\Python24\Scripts
Installing ipython.exe script to C:\Programs\Python24\Scripts
Installing pycolor-script.py script to C:\Programs\Python24\Scripts
Installing pycolor.exe script to C:\Programs\Python24\Scripts
Installing ipcontroller-script.py script to C:\Programs
\Python24\Scripts
Installing ipcontroller.exe script to C:\Programs\Python24\Scripts
Installing ipengine-script.py script to C:\Programs\Python24\Scripts
Installing ipengine.exe script to C:\Programs\Python24\Scripts

Installed c:\programs\python24\lib\site-packages\ipython-0.9-py2.4.egg
Processing dependencies for ipython==0.9
Finished processing dependencies for ipython==0.9


From emmanuelle.gouillart at normalesup.org  Sun Sep 14 10:02:17 2008
From: emmanuelle.gouillart at normalesup.org (Emmanuelle Gouillart)
Date: Sun, 14 Sep 2008 16:02:17 +0200
Subject: [SciPy-user] 2-D interpolation of irregularly spaced data
Message-ID: <20080914140217.GA917@phare.normalesup.org>

	Hello,

	I have an irregular 2-D mesh, and 1-D data measured at the
vertices of the mesh (the mesh is finer where the data vary more
rapidly), and I need to interpolate the data at other (also irregularly
spaced) points. To do so, I use the delaunay scikit and its
NNInterpolator which can take an iregular mesh. The problem is that I
cannot call the interpolator with irregularly spaced points, so that my
code is running very slowly now. Here is a minimal example of what I do
now (with regular grids and few points for clarity):

***
import scikits.delaunay as d

def evolve(positions, mesh, values):
    tri = d.Triangulation(mesh[0], mesh[1])
    interpolator = d.NNInterpolator(tri, values)
    return  array([interpolator(x,y) for (x,y) in positions.T]).ravel()

#Mesh    
X, Y = mgrid[-1:1:20j, -1:1:20j] #regular grid for clarity 
X = X.flatten()
Y = Y.flatten()
mesh = array([X,Y])
#Values
values = Y**2
#Positions
positions = mgrid[-0.5:0.5:20j, -0.5:0.5:20j]
positions = array([positions[0].flatten(), positions[1].flatten()])

new_values = evolve(positions, mesh, values)
***

Any hints about I could accelerate the interpolation? (Usually, I work
with meshes of size ~1.e4 and positions of size ~1.e6). Any help will be
very welcome!

Cheers,

Emmanuelle 


From gael.varoquaux at normalesup.org  Sun Sep 14 13:15:55 2008
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Sun, 14 Sep 2008 19:15:55 +0200
Subject: [SciPy-user] ipython install: egg not python 2.4 compatible
In-Reply-To: <a700e94c-2972-498a-8de8-5dd7b9a0cf7a@z66g2000hsc.googlegroups.com>
References: <a700e94c-2972-498a-8de8-5dd7b9a0cf7a@z66g2000hsc.googlegroups.com>
Message-ID: <20080914171555.GC12842@phare.normalesup.org>

On Sun, Sep 14, 2008 at 06:52:27AM -0700, joep wrote:
> Sorry for wrong group: I'm not subscribed to ipython-user:

This has been reported on the ipython mailing list. There will be a
bug-fix release to ipython to sort this out.

Ga?l


From chiefmurph at comcast.net  Sun Sep 14 21:44:25 2008
From: chiefmurph at comcast.net (Dan Murphy)
Date: Sun, 14 Sep 2008 18:44:25 -0700
Subject: [SciPy-user] Gaussian quadrature error
Message-ID: <0F8253EA348F49F6A3C3A0AEFE044866@GatewayLaptop>

I am trying out the integrate.quadrature function on the function f(x)=e**x
to the left of the y-axis. If the lower bound in not too negative, I get a
reasonable answer, but if the lower bound is too negative, I get 0.0 as the
value of the integral. Here is the code:

 

from scipy import *

 

def f(x):

    return e**x

 

integrate.quadrature(f,-10.0,0.0) # answer is (0.999954600065,
3.14148596026e-010)

 

but

 

integrate.quadrature(f,-1000.0,0.0) # yields (8.35116510531e-090,
8.35116510531e-090)

 

Note that 'val' and 'err' are equal. Is this a bug in quadrature?

 

Thanks.

Dan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080914/929ad341/attachment.html>

From robert.kern at gmail.com  Mon Sep 15 02:31:41 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 15 Sep 2008 01:31:41 -0500
Subject: [SciPy-user] 2-D interpolation of irregularly spaced data
In-Reply-To: <20080914140217.GA917@phare.normalesup.org>
References: <20080914140217.GA917@phare.normalesup.org>
Message-ID: <3d375d730809142331o65039239w86b23f5b881af61e@mail.gmail.com>

On Sun, Sep 14, 2008 at 09:02, Emmanuelle Gouillart
<emmanuelle.gouillart at normalesup.org> wrote:
>        Hello,
>
>        I have an irregular 2-D mesh, and 1-D data measured at the
> vertices of the mesh (the mesh is finer where the data vary more
> rapidly), and I need to interpolate the data at other (also irregularly
> spaced) points. To do so, I use the delaunay scikit and its
> NNInterpolator which can take an iregular mesh. The problem is that I
> cannot call the interpolator with irregularly spaced points, so that my
> code is running very slowly now. Here is a minimal example of what I do
> now (with regular grids and few points for clarity):
>
> ***
> import scikits.delaunay as d
>
> def evolve(positions, mesh, values):
>    tri = d.Triangulation(mesh[0], mesh[1])
>    interpolator = d.NNInterpolator(tri, values)
>    return  array([interpolator(x,y) for (x,y) in positions.T]).ravel()

NNInterpolator.__call__() can take arrays, not just scalars. For the
greatest efficiency, try to make sure adjacent points are close to
each other.

-- 
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 emmanuelle.gouillart at normalesup.org  Mon Sep 15 07:17:31 2008
From: emmanuelle.gouillart at normalesup.org (Emmanuelle Gouillart)
Date: Mon, 15 Sep 2008 13:17:31 +0200 (CEST)
Subject: [SciPy-user] 2-D interpolation of irregularly spaced data
In-Reply-To: <3d375d730809142331o65039239w86b23f5b881af61e@mail.gmail.com>
References: <20080914140217.GA917@phare.normalesup.org>
	<3d375d730809142331o65039239w86b23f5b881af61e@mail.gmail.com>
Message-ID: <46468.195.68.31.231.1221477451.squirrel@www.normalesup.org>

Thank you, it works really fast with arrays! I don't know why I was
convinced NNInterpolator.__call__() could only take arrays with regularly
spaced values...

Thanks a lot,
Emmanuelle


> NNInterpolator.__call__() can take arrays, not just scalars. For the
> greatest efficiency, try to make sure adjacent points are close to
> each other.
>
> --
> Robert Kern
>
> "I have come to believe that the whole world is an enigma, a harmless
> enigma that is made terrible by our own mad attempt to interpret it as
> though it had an underlying truth."
>  -- Umberto Eco
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>




From cyril.giraudon at free.fr  Mon Sep 15 08:49:14 2008
From: cyril.giraudon at free.fr (cyril giraudon)
Date: Mon, 15 Sep 2008 14:49:14 +0200
Subject: [SciPy-user] butterworth filter
In-Reply-To: <48C67C16.9030701@free.fr>
References: <48C67C16.9030701@free.fr>
Message-ID: <48CE59CA.5070609@free.fr>

I didn't understand abscissa differences.
in fact, x (normalized frequency) are devided by pi.

However, in low frequency, scipy trunsfer function is very low.
At 0.2, scipy : -250dB, matlab : -120dB.

Is there any explanation ?

Thanks a lot,

Cyril.





-------------- next part --------------
A non-text attachment was scrubbed...
Name: butter_2_28.gif
Type: image/gif
Size: 19261 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080915/fb80414b/attachment.gif>

From dmitrey.kroshko at scipy.org  Mon Sep 15 09:18:17 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Mon, 15 Sep 2008 16:18:17 +0300
Subject: [SciPy-user] [numerical optimization] OpenOpt release v 0.19
Message-ID: <48CE6099.1060409@scipy.org>

Hi all,

I'm glad to inform you about new OpenOpt release: v 0.19.
http://openopt.blogspot.com/2008/09/openopt-release-019.html

OpenOpt is free (license: BSD) numerical optimization with lots of 3rd 
party solvers connected (some are C or Fortran written) and some own 
ones, graphic output and lots of other numerical optimization "MUST 
HAVE" features.

For more details see OpenOpt's
- homepage: http://scipy.org/scipy/scikits/wiki/OpenOpt
- foreword: http://scipy.org/scipy/scikits/wiki/OOForeword

Regards, Dmitrey.


From robert.kern at gmail.com  Mon Sep 15 11:10:16 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 15 Sep 2008 10:10:16 -0500
Subject: [SciPy-user] 2-D interpolation of irregularly spaced data
In-Reply-To: <46468.195.68.31.231.1221477451.squirrel@www.normalesup.org>
References: <20080914140217.GA917@phare.normalesup.org>
	<3d375d730809142331o65039239w86b23f5b881af61e@mail.gmail.com>
	<46468.195.68.31.231.1221477451.squirrel@www.normalesup.org>
Message-ID: <3d375d730809150810y6b5fdd46lca695a9fbcac4b52@mail.gmail.com>

On Mon, Sep 15, 2008 at 06:17, Emmanuelle Gouillart
<emmanuelle.gouillart at normalesup.org> wrote:
> Thank you, it works really fast with arrays! I don't know why I was
> convinced NNInterpolator.__call__() could only take arrays with regularly
> spaced values...

The docstring is out of date and suggests that only regular grids are supported.

-- 
Robert Kern

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


From anand.prabhakar.patil at gmail.com  Mon Sep 15 11:55:36 2008
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Mon, 15 Sep 2008 16:55:36 +0100
Subject: [SciPy-user] Breaking up an array of indices into contiguous/
	strided chunks
Message-ID: <2bc7a5a50809150855l1c86c659l9bd2e4528988c239@mail.gmail.com>

Hi all,
I have an array of indices that I want to use as a slice, but cvxopt sparse
matrices don't seem to take fancy slices so I need to take the slice
manually. If I just do it element-by-element in a for-loop it's extremely
slow. I know that the array of indices is in ascending order, but that's
all. Does anyone know an easy way to break it up into strided chunks that
are as big as possible?

Thanks,
Anand
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080915/4e895674/attachment.html>

From mathieu.dubois at limsi.fr  Mon Sep 15 12:13:36 2008
From: mathieu.dubois at limsi.fr (Mathieu Dubois)
Date: Mon, 15 Sep 2008 18:13:36 +0200
Subject: [SciPy-user] How to give a name to a figure?
Message-ID: <48CE89B0.5010901@limsi.fr>

Hi,

I'm a beginner in scipy and I have a small problem with figures. Let me 
explain.

I have to plot complicated data so I have a lot of figures. I have set 
title and axes names.

My problem is that the windows are titled with things like 'Figure 1', 
'Figure 2' etc. Is it possible to set this name to something more 
understable? Apparently title() changes the name of the plots inside the 
figure.

My goal is to save (with savefig()) them with a nice name say 
'title.png' (where title would be replaced by the title of my figure).

Any help would be appreciated.

Thanks in advance,
Mathieu


From robert.kern at gmail.com  Mon Sep 15 12:31:16 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 15 Sep 2008 11:31:16 -0500
Subject: [SciPy-user] How to give a name to a figure?
In-Reply-To: <48CE89B0.5010901@limsi.fr>
References: <48CE89B0.5010901@limsi.fr>
Message-ID: <3d375d730809150931p27192044kb5d85985a30ec955@mail.gmail.com>

On Mon, Sep 15, 2008 at 11:13, Mathieu Dubois <mathieu.dubois at limsi.fr> wrote:
> Hi,
>
> I'm a beginner in scipy and I have a small problem with figures. Let me
> explain.

However, this is a question about matplotlib, not scipy. The mailing
list you want is over here:

  https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

-- 
Robert Kern

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


From anand.prabhakar.patil at gmail.com  Mon Sep 15 13:03:02 2008
From: anand.prabhakar.patil at gmail.com (Anand Patil)
Date: Mon, 15 Sep 2008 18:03:02 +0100
Subject: [SciPy-user] Breaking up an array of indices into contiguous/
	strided chunks
In-Reply-To: <2bc7a5a50809150855l1c86c659l9bd2e4528988c239@mail.gmail.com>
References: <2bc7a5a50809150855l1c86c659l9bd2e4528988c239@mail.gmail.com>
Message-ID: <2bc7a5a50809151003s2d08d735h896cd6a6db6ce856@mail.gmail.com>

On Mon, Sep 15, 2008 at 4:55 PM, Anand Patil <
anand.prabhakar.patil at gmail.com> wrote:

> Hi all,
> I have an array of indices that I want to use as a slice, but cvxopt sparse
> matrices don't seem to take fancy slices so I need to take the slice
> manually. If I just do it element-by-element in a for-loop it's extremely
> slow. I know that the array of indices is in ascending order, but that's
> all. Does anyone know an easy way to break it up into strided chunks that
> are as big as possible?
>

I didn't ask the question very well. I'm trying to do

x[i1] = y[i2],

where i1 and i2 are different but they're both in ascending order and I have
reason to believe that they're both regular for long intervals, though the
strides will usually be different. I need to break them up into matching,
strided chunks.

The problem gets weird when, for example,

i1 = [1,2,4,5,7,8,10...]
i2 = [1,2,3,4,5,6,7...]

that is when the stride of i1 'alternates'. In this case the copy could be
done efficiently using

s1 = [slice(1,?,3), slice(2,?,3)]
s2 = [slice(1,?,2), slice(2,?,2)]

for k in (1,2):
   x[s1[k]] = y[s2[k]]

and it seems like some code to do this kind of pattern matching would
already be available. Maybe numpy even does it?

Thanks,
Anand
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080915/553acb43/attachment.html>

From bnuttall at uky.edu  Mon Sep 15 12:28:35 2008
From: bnuttall at uky.edu (Nuttall, Brandon C)
Date: Mon, 15 Sep 2008 12:28:35 -0400
Subject: [SciPy-user] How to give a name to a figure?
In-Reply-To: <48CE89B0.5010901@limsi.fr>
References: <48CE89B0.5010901@limsi.fr>
Message-ID: <F5BC85B8534D524DB972D2B2731CA53F29CD7B846E@EX7FM01.ad.uky.edu>

Mathieu,

Its pretty easy. The statements below do what you I want:

(of course, you have to import pylab)

>>> path = 'c:\\documents and settings\\bnuttall\\desktop\\adair\\'
>>> picfile = '%sR%s.png' % (path,str(wellid).rjust(7,'0'))

...snip... (a bunch of statements constructing the figure)

>>> pylab.savefig(picfile)

In my case, the wellid is a serial integer that uniquely identifies the data set. I have coded similar statements to alter the path name systematically so that related output gets grouped in folders.

Brandon Nuttall

-----Original Message-----
From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org] On Behalf Of Mathieu Dubois
Sent: Monday, September 15, 2008 12:14 PM
To: scipy-user at scipy.org
Subject: [SciPy-user] How to give a name to a figure?

Hi,

I'm a beginner in scipy and I have a small problem with figures. Let me
explain.

I have to plot complicated data so I have a lot of figures. I have set
title and axes names.

My problem is that the windows are titled with things like 'Figure 1',
'Figure 2' etc. Is it possible to set this name to something more
understable? Apparently title() changes the name of the plots inside the
figure.

My goal is to save (with savefig()) them with a nice name say
'title.png' (where title would be replaced by the title of my figure).

Any help would be appreciated.

Thanks in advance,
Mathieu
_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org
http://projects.scipy.org/mailman/listinfo/scipy-user


From ndbecker2 at gmail.com  Mon Sep 15 13:50:31 2008
From: ndbecker2 at gmail.com (Neal Becker)
Date: Mon, 15 Sep 2008 13:50:31 -0400
Subject: [SciPy-user] [numerical optimization] OpenOpt release v 0.19
References: <48CE6099.1060409@scipy.org>
Message-ID: <gam797$lj6$1@ger.gmane.org>

Thanks.  One buglet:

python setup.py build
sudo python setup.py install

leaves files in build dir owned by root:
 scikits.openopt.egg-info/



From f.drossaert at googlemail.com  Mon Sep 15 13:45:26 2008
From: f.drossaert at googlemail.com (Francis Drossaert)
Date: Mon, 15 Sep 2008 18:45:26 +0100
Subject: [SciPy-user] Help: building scipy+numpy locally on x86_64 running
	Centos
Message-ID: <2c758b440809151045m4eb35129v6c1eecc0e0c1ed5d@mail.gmail.com>

Hi everybody,

I am trying to install python2.5/scipy/numpy/sympy/matplotlib locally,
because of various reasons. I am not root.

At home I am using Ubuntu and I am root and everything works as it should.

So far I managed to make python2.5, sympy and ipython to run fine on
my computer at work, but I am struggling to make numpy+scipy since I
am trying to install it locally. The problem is likely the
lapack+atlas build. I have followed the build by hand instructions on
http://www.scipy.org/Installing_SciPy/Linux. I am not getting any
errors during the build and install.

However after installing I tried to import numpy in python and
received to following error:  ImportError:
/users/francisd/local/lib/python2.5/site-packages/numpy/linalg/lapack_lite.so:
undefined symbol: zgesdd_

Googling this error, it seems that the error is caused by having a
wrong lapack version. I am running Centos 5 (basically Red Hat
Enterprise Linux 5) on a x86_64 machine.

Does anybody know how I can get the right version, by changing some
flags, or whatever? FYI I have added the -m64 flag for the lapack
build, but no change.

Cheers,

Francis


From robert.kern at gmail.com  Mon Sep 15 13:55:38 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 15 Sep 2008 12:55:38 -0500
Subject: [SciPy-user] [numerical optimization] OpenOpt release v 0.19
In-Reply-To: <gam797$lj6$1@ger.gmane.org>
References: <48CE6099.1060409@scipy.org> <gam797$lj6$1@ger.gmane.org>
Message-ID: <3d375d730809151055u2a6ee1b4k87090a3381392d9f@mail.gmail.com>

On Mon, Sep 15, 2008 at 12:50, Neal Becker <ndbecker2 at gmail.com> wrote:
> Thanks.  One buglet:
>
> python setup.py build
> sudo python setup.py install
>
> leaves files in build dir owned by root:
>  scikits.openopt.egg-info/

That's not OpenOpt's fault. If you wish to avoid it, do

  $ python setup.py build egg_info
  $ sudo python setup.py install

-- 
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 george.dahl at gmail.com  Mon Sep 15 14:01:26 2008
From: george.dahl at gmail.com (George Dahl)
Date: Mon, 15 Sep 2008 14:01:26 -0400
Subject: [SciPy-user] trouble building scipy on mac os 10.5 intel
In-Reply-To: <6b2e0e10809151059o359268acm1760fa62f42cb2a3@mail.gmail.com>
References: <6b2e0e10809070018m1dbfcc6esbe13ac5e92e22e82@mail.gmail.com>
	<6b2e0e10809151059o359268acm1760fa62f42cb2a3@mail.gmail.com>
Message-ID: <6b2e0e10809151101k433af21bv44f3609a3e92b8e3@mail.gmail.com>

Hi everyone.  I get the following result when I run python setup.py
build for scipy 0.6.0.
I have numpy and it works in python2.5, which is the version of python
I want to use scipy from.  I have gcc 4.0.1 and gfortran 4.2.1.

I have pasted some of the output of python setup.py build below.  I
don't really know what I am doing, but I would really like to get
scipy working and I would appreciate any help anyone can give me!  I
looked around in the archives with google a bit, but nothing seemed to
deal with my situation exactly, hopefully I didn't miss anything.
- George

scipy-0.6.0$ python setup.py build
.
.
.
  bnrm2,resid,info = dstoptest2(r,b,bnrm2,tol,info)
Constructing wrapper function "cstoptest2"...
 bnrm2,resid,info = cstoptest2(r,b,bnrm2,tol,info)
Constructing wrapper function "zstoptest2"...
 bnrm2,resid,info = zstoptest2(r,b,bnrm2,tol,info)
Wrote C/API module "_iterative" to file
"build/src.macosx-10.3-fat-2.5/build/src.macosx-10.3-fat-2.5/scipy/linalg/iterative/_iterativemodule.c"
 adding 'build/src.macosx-10.3-fat-2.5/fortranobject.c' to sources.
 adding 'build/src.macosx-10.3-fat-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
creating build/src.macosx-10.3-fat-2.5/scipy/linsolve
creating build/src.macosx-10.3-fat-2.5/scipy/linsolve/umfpack
 adding 'scipy/linsolve/umfpack/umfpack.i' to sources.
swig: scipy/linsolve/umfpack/umfpack.i
swig -python -o
build/src.macosx-10.3-fat-2.5/scipy/linsolve/umfpack/_umfpack_wrap.c
-outdir build/src.macosx-10.3-fat-2.5/scipy/linsolve/umfpack
scipy/linsolve/umfpack/umfpack.i
scipy/linsolve/umfpack/umfpack.i:192: Error: Unable to find 'umfpack.h'
scipy/linsolve/umfpack/umfpack.i:193: Error: Unable to find 'umfpack_solve.h'
scipy/linsolve/umfpack/umfpack.i:194: Error: Unable to find 'umfpack_defaults.h'
scipy/linsolve/umfpack/umfpack.i:195: Error: Unable to find
'umfpack_triplet_to_col.h'
scipy/linsolve/umfpack/umfpack.i:196: Error: Unable to find
'umfpack_col_to_triplet.h'
scipy/linsolve/umfpack/umfpack.i:197: Error: Unable to find
'umfpack_transpose.h'
scipy/linsolve/umfpack/umfpack.i:198: Error: Unable to find 'umfpack_scale.h'
scipy/linsolve/umfpack/umfpack.i:200: Error: Unable to find
'umfpack_report_symbolic.h'
scipy/linsolve/umfpack/umfpack.i:201: Error: Unable to find
'umfpack_report_numeric.h'
scipy/linsolve/umfpack/umfpack.i:202: Error: Unable to find
'umfpack_report_info.h'
scipy/linsolve/umfpack/umfpack.i:203: Error: Unable to find
'umfpack_report_control.h'
scipy/linsolve/umfpack/umfpack.i:215: Error: Unable to find 'umfpack_symbolic.h'
scipy/linsolve/umfpack/umfpack.i:216: Error: Unable to find 'umfpack_numeric.h'
scipy/linsolve/umfpack/umfpack.i:225: Error: Unable to find
'umfpack_free_symbolic.h'
scipy/linsolve/umfpack/umfpack.i:226: Error: Unable to find
'umfpack_free_numeric.h'
scipy/linsolve/umfpack/umfpack.i:248: Error: Unable to find 'umfpack_get_lunz.h'
scipy/linsolve/umfpack/umfpack.i:272: Error: Unable to find
'umfpack_get_numeric.h'
error: command 'swig' failed with exit status 1


From mathieu.dubois at limsi.fr  Mon Sep 15 14:38:22 2008
From: mathieu.dubois at limsi.fr (Mathieu Dubois)
Date: Mon, 15 Sep 2008 20:38:22 +0200
Subject: [SciPy-user] How to give a name to a figure?
In-Reply-To: <F5BC85B8534D524DB972D2B2731CA53F29CD7B846E@EX7FM01.ad.uky.edu>
References: <48CE89B0.5010901@limsi.fr>
	<F5BC85B8534D524DB972D2B2731CA53F29CD7B846E@EX7FM01.ad.uky.edu>
Message-ID: <48CEAB9E.5060403@limsi.fr>

Hi,

Thanks for your help but my problem is not to give a name to the file 
but to give a name to the figure itself.

The reason why I want this is that I have all my figure handles in a 
list and I want to do a loop like:
for fig in fig_list
    savefig(fig, fig.title, format='png')

The question was asked to the matplotlib users list some months ago:
http://sourceforge.net/mailarchive/message.php?msg_id=a7f1ef730709101012o20abd37aj116e100d9b105d52%40mail.gmail.com

As Robert Kern pointed out this question is matplotlib related and I 
don't want to pollute scipy list so I will continue discussion on 
Matplotlib-users list.

Thanks again for your help,
Mathieu

Nuttall, Brandon C wrote:
> Mathieu,
>
> Its pretty easy. The statements below do what you I want:
>
> (of course, you have to import pylab)
>
>   
>>>> path = 'c:\\documents and settings\\bnuttall\\desktop\\adair\\'
>>>> picfile = '%sR%s.png' % (path,str(wellid).rjust(7,'0'))
>>>>         
>
> ...snip... (a bunch of statements constructing the figure)
>
>   
>>>> pylab.savefig(picfile)
>>>>         
>
> In my case, the wellid is a serial integer that uniquely identifies the data set. I have coded similar statements to alter the path name systematically so that related output gets grouped in folders.
>
> Brandon Nuttall
>
> -----Original Message-----
> From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org] On Behalf Of Mathieu Dubois
> Sent: Monday, September 15, 2008 12:14 PM
> To: scipy-user at scipy.org
> Subject: [SciPy-user] How to give a name to a figure?
>
> Hi,
>
> I'm a beginner in scipy and I have a small problem with figures. Let me
> explain.
>
> I have to plot complicated data so I have a lot of figures. I have set
> title and axes names.
>
> My problem is that the windows are titled with things like 'Figure 1',
> 'Figure 2' etc. Is it possible to set this name to something more
> understable? Apparently title() changes the name of the plots inside the
> figure.
>
> My goal is to save (with savefig()) them with a nice name say
> 'title.png' (where title would be replaced by the title of my figure).
>
> Any help would be appreciated.
>
> Thanks in advance,
> Mathieu
> _______________________________________________
> 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 ndbecker2 at gmail.com  Mon Sep 15 14:54:52 2008
From: ndbecker2 at gmail.com (Neal Becker)
Date: Mon, 15 Sep 2008 14:54:52 -0400
Subject: [SciPy-user] [numerical optimization] OpenOpt release v 0.19
References: <48CE6099.1060409@scipy.org> <gam797$lj6$1@ger.gmane.org>
	<3d375d730809151055u2a6ee1b4k87090a3381392d9f@mail.gmail.com>
Message-ID: <gamb1t$3sr$1@ger.gmane.org>

Robert Kern wrote:

> On Mon, Sep 15, 2008 at 12:50, Neal Becker <ndbecker2 at gmail.com> wrote:
>> Thanks.  One buglet:
>>
>> python setup.py build
>> sudo python setup.py install
>>
>> leaves files in build dir owned by root:
>>  scikits.openopt.egg-info/
> 
> That's not OpenOpt's fault. If you wish to avoid it, do
> 
>   $ python setup.py build egg_info
>   $ sudo python setup.py install
> 

OK, thanks.  Please add to install instructions.



From spmcinerney at hotmail.com  Mon Sep 15 18:45:11 2008
From: spmcinerney at hotmail.com (Stephen McInerney)
Date: Mon, 15 Sep 2008 15:45:11 -0700
Subject: [SciPy-user] SciPy 08 conference videos?
In-Reply-To: <mailman.11.1221498006.17129.scipy-user@scipy.org>
References: <mailman.11.1221498006.17129.scipy-user@scipy.org>
Message-ID: <BAY111-W4376BC42186AD93C342DFA0520@phx.gbl>


Enthought folks,
 
the slides are up at http://conference.scipy.org/
but can you also put up the videos (unedited is fine, for starters)
 
Thanks,
Stephen
_________________________________________________________________
Get more out of the Web. Learn 10 hidden secrets of Windows Live.
http://windowslive.com/connect/post/jamiethomson.spaces.live.com-Blog-cns!550F681DAD532637!5295.entry?ocid=TXT_TAGLM_WL_domore_092008
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080915/7870c167/attachment.html>

From lanceboyle at qwest.net  Tue Sep 16 01:00:40 2008
From: lanceboyle at qwest.net (Jerry)
Date: Mon, 15 Sep 2008 22:00:40 -0700
Subject: [SciPy-user] butterworth filter
In-Reply-To: <c5b438120809121418u3ca83865o521271be3c67f70c@mail.gmail.com>
References: <48C67C16.9030701@free.fr>
	<c5b438120809121418u3ca83865o521271be3c67f70c@mail.gmail.com>
Message-ID: <B4CE4E9C-FAAB-4311-985B-0A8B5EA171BD@qwest.net>

Just an innocent bystander here, but that plot looks wrong. Could you  
please re-submit it with decibels on the vertical axis against log  
frequency? That is the normal way of viewing filter frequency responses.
Jerry

On Sep 12, 2008, at 2:18 PM, Ryan Krauss wrote:

> So, attached is the plot I get from scipy.  It is a high pass  
> filter.  It seems reasonable.  What does the curve look like from  
> Matlab.
>
> Ryan
>
> On Tue, Sep 9, 2008 at 8:37 AM, cyril giraudon  
> <cyril.giraudon at free.fr> wrote:
> Hi,
>
> I use scipy 0.6.0 and i try to reproduce the plot of the matlab butter
> function web documentation (first response for a google request  
> "matlab
> butter example").
>
> The matlab code is :
>
> [z,p,k] = butter(9,300/500,'high');
> [sos,g] = zp2sos(z,p,k);             % Convert to SOS form
> Hd = dfilt.df2tsos(sos,g);   % Create a dfilt object
> h = fvtool(Hd);              % Plot magnitude response
> set(h,'Analysis','freq')             % Display frequency response
>
>
> In scipy, I write :
>
> from scipy.signal import butter, freqz
>
> from pylab import show, grid, log, plot
>
> b, a = butter(9, 300./500., 'high')
>
> fi = freqz(b, a)
>
> plot(fi[0], 20*log(abs(fi[1])))
>
> grid()
>
> show()
>
>
>
> Why are the two filters not the same ?
>
>
> Thanks a lot,
>
> Cyril.
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
> <butter_scipy.png>_______________________________________________
> 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/20080915/6dc8ff43/attachment.html>

From loniedavid at gmail.com  Tue Sep 16 13:06:32 2008
From: loniedavid at gmail.com (David Lonie)
Date: Tue, 16 Sep 2008 12:06:32 -0500
Subject: [SciPy-user] Error in nonlinear least squares fit analysis
Message-ID: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>

I got some help here earlier about finding a function to fit a
function to some exponentially increasing data. I have a few
questions:

a) fmin vs. leastsq:
The method I wrote ended up using the fmin() function to minimize the
error vector. What is the difference between fmin and leastsq? Is
there an advantage to using either?

b) Error in the parameters:
I'd like to know the precision that the fitted parameters are good to.
Basically, I'd like to know that b = 3.456 +/- 0.003 instead of just b
= 3.456.
leastsq can return a Jacobian matrix -- will pulling out the diagonal
elements of this matrix give me the results I want? Or is there a
better way?

TIA,

Dave


From s.mientki at ru.nl  Tue Sep 16 13:25:58 2008
From: s.mientki at ru.nl (Stef Mientki)
Date: Tue, 16 Sep 2008 19:25:58 +0200
Subject: [SciPy-user] butterworth filter
In-Reply-To: <48CE59CA.5070609@free.fr>
References: <48C67C16.9030701@free.fr> <48CE59CA.5070609@free.fr>
Message-ID: <48CFEC26.40505@ru.nl>



cyril giraudon wrote:
> I didn't understand abscissa differences.
> in fact, x (normalized frequency) are devided by pi.
>
> However, in low frequency, scipy trunsfer function is very low.
> At 0.2, scipy : -250dB, matlab : -120dB.
As an engineer I would say -120 dB equals -250 dB, so the graphs are equal.
I really would like to see your lab, if your interested -250 dB ;-)
In Matlab you probably never get to know how the frequency response is 
calculated,
although presumable it's just a copy of  "Numerical recipes".
At  -120 dB, the rounding results might become relevant.
It would be interesting to compare the calculated filter coefficients,
or the zero/pole values,
and maybe even put the Matlab coefficients into Scipy and vise versa.

cheers,
Stef


>
> Is there any explanation ?
>
> Thanks a lot,
>
> Cyril.
>
>
>
>
>
>
> ------------------------------------------------------------------------
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   


From nwagner at iam.uni-stuttgart.de  Tue Sep 16 15:04:44 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Tue, 16 Sep 2008 21:04:44 +0200
Subject: [SciPy-user] Rombouts algorithm for calculating the characteristic
 polynomial
Message-ID: <web-111378903@uni-stuttgart.de>

Hi all,

Has someone implemented Rombouts algorithm in Python ?

Details are available at
http://arxiv.org/pdf/math/9804133v1

Cheers,

            Nils

  


From david at ar.media.kyoto-u.ac.jp  Tue Sep 16 22:53:57 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 17 Sep 2008 11:53:57 +0900
Subject: [SciPy-user] butterworth filter
In-Reply-To: <48CFEC26.40505@ru.nl>
References: <48C67C16.9030701@free.fr> <48CE59CA.5070609@free.fr>
	<48CFEC26.40505@ru.nl>
Message-ID: <48D07145.2010605@ar.media.kyoto-u.ac.jp>

Stef Mientki wrote:
> As an engineer I would say -120 dB equals -250 dB, so the graphs are equal.
> I really would like to see your lab, if your interested -250 dB ;-)
> In Matlab you probably never get to know how the frequency response is 
> calculated,

That's not entirely accurate. A lot of matlab functions are implemented
in matlab itself. That's the case for butterworth AFAICS (butterworth is
a big .m file, suggesting the implementation itself is in matlab). I
avoid reading the actual code itself, though, for licensing issues,
because I work a lot on scipy/numpy. But if you don't intend to write
this code for numpy/scipy, I guess you can read it.

> although presumable it's just a copy of  "Numerical recipes".

IIRC, there is no signal processing code in Numerical recipes. For
Butterworth, it should not be too difficult to check, though, because it
is a direct implementation of the analog domain filters. That's
certainly one of the easiest IIR to implement for anyone familiar with
digital signal processing (bilinear transform and co to go into digital
domain).

Looking at the code, for butterworth, I think the scipy implementation
may be a bit too naive in some corner cases. IIRC, you would be better
implementing a N order filter by cascading 2n order filters. In a former
life, I was really into synthesizer and digital sound processing, and I
know that it mattered in fixed point and even floating point (32 bits)
implementations, because for synthesizer, you like the corner cases (to
make boom boom and make you ear bleed in clubs).

cheers,

David


From john.grosspietsch at gmail.com  Wed Sep 17 00:14:44 2008
From: john.grosspietsch at gmail.com (John Grosspietsch)
Date: Wed, 17 Sep 2008 04:14:44 +0000 (UTC)
Subject: [SciPy-user] butterworth filter
References: <48C67C16.9030701@free.fr> <48CE59CA.5070609@free.fr>
Message-ID: <loom.20080917T041019-207@post.gmane.org>

cyril giraudon <cyril.giraudon <at> free.fr> writes:

> 
> I didn't understand abscissa differences.
> in fact, x (normalized frequency) are devided by pi.
> 
> However, in low frequency, scipy trunsfer function is very low.
> At 0.2, scipy : -250dB, matlab : -120dB.
> 
> Is there any explanation ?
> 
> Thanks a lot,
> 
> Cyril.
> 

You need log10() not log() to calculate the filter response in decibels.  The
response is a little distorted near the corner frequency because, I assume, of
the bilinear transformation used to map analog filter poles to digital poles. 

John



From s.mientki at ru.nl  Wed Sep 17 03:59:26 2008
From: s.mientki at ru.nl (Stef Mientki)
Date: Wed, 17 Sep 2008 09:59:26 +0200
Subject: [SciPy-user] butterworth filter
In-Reply-To: <48D07145.2010605@ar.media.kyoto-u.ac.jp>
References: <48C67C16.9030701@free.fr> <48CE59CA.5070609@free.fr>
	<48CFEC26.40505@ru.nl> <48D07145.2010605@ar.media.kyoto-u.ac.jp>
Message-ID: <48D0B8DE.5010006@ru.nl>

thanks David,

for correcting me.
Sorry, I think I generalized a problem I once had with Matlab too much.
But anyway such a statement reveals some of the essential details ;-)

cheers,
Stef

David Cournapeau wrote:
> Stef Mientki wrote:
>   
>> As an engineer I would say -120 dB equals -250 dB, so the graphs are equal.
>> I really would like to see your lab, if your interested -250 dB ;-)
>> In Matlab you probably never get to know how the frequency response is 
>> calculated,
>>     
>
> That's not entirely accurate. A lot of matlab functions are implemented
> in matlab itself. That's the case for butterworth AFAICS (butterworth is
> a big .m file, suggesting the implementation itself is in matlab). I
> avoid reading the actual code itself, though, for licensing issues,
> because I work a lot on scipy/numpy. But if you don't intend to write
> this code for numpy/scipy, I guess you can read it.
>
>   
>> although presumable it's just a copy of  "Numerical recipes".
>>     
>
> IIRC, there is no signal processing code in Numerical recipes. For
> Butterworth, it should not be too difficult to check, though, because it
> is a direct implementation of the analog domain filters. That's
> certainly one of the easiest IIR to implement for anyone familiar with
> digital signal processing (bilinear transform and co to go into digital
> domain).
>
> Looking at the code, for butterworth, I think the scipy implementation
> may be a bit too naive in some corner cases. IIRC, you would be better
> implementing a N order filter by cascading 2n order filters. In a former
> life, I was really into synthesizer and digital sound processing, and I
> know that it mattered in fixed point and even floating point (32 bits)
> implementations, because for synthesizer, you like the corner cases (to
> make boom boom and make you ear bleed in clubs).
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>   


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 c.j.lee at tnw.utwente.nl  Wed Sep 17 04:20:03 2008
From: c.j.lee at tnw.utwente.nl (Chris Lee)
Date: Wed, 17 Sep 2008 10:20:03 +0200
Subject: [SciPy-user] singular value decomposition
Message-ID: <9468F765-C9A0-41C6-9632-B8F8B3A88ADE@tnw.utwente.nl>

Hi All,

I have code that needs to repeatedly perform an svd on a matrix (2304,  
2304) shape. It seems to take forever to do (I haven't timed it, but  
approximately two hours passed on the first trial (without completion)  
and I am now running a second to make sure nothing went wrong.

Is this because I am not linked to an optimized BLAS/LAPACK library or  
something else?

I have access to mkl. Is it possible to link an already installed  
scipy to mkl, or would I have to reinstall?

Cheers
Chris

***************************************************
Chris Lee
Laser Physics and Nonlinear Optics Group
MESA+ Research Institute for Nanotechnology
University of Twente
Phone: ++31 (0)53 489 3968
fax: ++31 (0)53 489 1102
***************************************************



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080917/cb596b05/attachment.html>

From robert.kern at gmail.com  Wed Sep 17 04:25:31 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 17 Sep 2008 03:25:31 -0500
Subject: [SciPy-user] singular value decomposition
In-Reply-To: <9468F765-C9A0-41C6-9632-B8F8B3A88ADE@tnw.utwente.nl>
References: <9468F765-C9A0-41C6-9632-B8F8B3A88ADE@tnw.utwente.nl>
Message-ID: <3d375d730809170125y3d936573sa02de53a9b258cea@mail.gmail.com>

On Wed, Sep 17, 2008 at 03:20, Chris Lee <c.j.lee at tnw.utwente.nl> wrote:
> Hi All,
> I have code that needs to repeatedly perform an svd on a matrix (2304, 2304)
> shape. It seems to take forever to do (I haven't timed it, but approximately
> two hours passed on the first trial (without completion) and I am now
> running a second to make sure nothing went wrong.
> Is this because I am not linked to an optimized BLAS/LAPACK library or
> something else?

Possibly. The same sized problem takes 40s for me using OS X's builtin ATLAS.

> I have access to mkl. Is it possible to link an already installed scipy to
> mkl, or would I have to reinstall?

You would have to reinstall.

-- 
Robert Kern

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


From david at ar.media.kyoto-u.ac.jp  Wed Sep 17 04:12:13 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 17 Sep 2008 17:12:13 +0900
Subject: [SciPy-user] singular value decomposition
In-Reply-To: <9468F765-C9A0-41C6-9632-B8F8B3A88ADE@tnw.utwente.nl>
References: <9468F765-C9A0-41C6-9632-B8F8B3A88ADE@tnw.utwente.nl>
Message-ID: <48D0BBDD.7040605@ar.media.kyoto-u.ac.jp>

Chris Lee wrote:
> Hi All,
>
> I have code that needs to repeatedly perform an svd on a matrix (2304,
> 2304) shape. It seems to take forever to do (I haven't timed it, but
> approximately two hours passed on the first trial (without completion)
> and I am now running a second to make sure nothing went wrong.
>
> Is this because I am not linked to an optimized BLAS/LAPACK library or
> something else?

Have you tested your scipy atlas ? 2 hours sounds awfully long, and this
may be due to some bugs caused during a wrong build (for example, errors
in the masrhalling of arguments at the C/Fortran interface),

cheers,

David


From c.j.lee at tnw.utwente.nl  Wed Sep 17 04:32:52 2008
From: c.j.lee at tnw.utwente.nl (Chris Lee)
Date: Wed, 17 Sep 2008 10:32:52 +0200
Subject: [SciPy-user] singular value decomposition
In-Reply-To: <48D0BBDD.7040605@ar.media.kyoto-u.ac.jp>
References: <9468F765-C9A0-41C6-9632-B8F8B3A88ADE@tnw.utwente.nl>
	<48D0BBDD.7040605@ar.media.kyoto-u.ac.jp>
Message-ID: <4CDB6AD0-C886-4264-878E-4E55C9F3CFEB@tnw.utwente.nl>

Thanks David and Robert, at least I am now sure something is wrong. I  
will see what libraries I am linked to and see what can be done to fix  
the problem.

Cheers
Chris
On Sep 17, 2008, at 10:12 AM, David Cournapeau wrote:

> Chris Lee wrote:
>> Hi All,
>>
>> I have code that needs to repeatedly perform an svd on a matrix  
>> (2304,
>> 2304) shape. It seems to take forever to do (I haven't timed it, but
>> approximately two hours passed on the first trial (without  
>> completion)
>> and I am now running a second to make sure nothing went wrong.
>>
>> Is this because I am not linked to an optimized BLAS/LAPACK library  
>> or
>> something else?
>
> Have you tested your scipy atlas ? 2 hours sounds awfully long, and  
> this
> may be due to some bugs caused during a wrong build (for example,  
> errors
> in the masrhalling of arguments at the C/Fortran interface),
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

***************************************************
Chris Lee
Laser Physics and Nonlinear Optics Group
MESA+ Research Institute for Nanotechnology
University of Twente
Phone: ++31 (0)53 489 3968
fax: ++31 (0)53 489 1102
***************************************************





From c.j.lee at tnw.utwente.nl  Wed Sep 17 05:33:39 2008
From: c.j.lee at tnw.utwente.nl (Chris Lee)
Date: Wed, 17 Sep 2008 11:33:39 +0200
Subject: [SciPy-user] singular value decomposition
In-Reply-To: <48D0BBDD.7040605@ar.media.kyoto-u.ac.jp>
References: <9468F765-C9A0-41C6-9632-B8F8B3A88ADE@tnw.utwente.nl>
	<48D0BBDD.7040605@ar.media.kyoto-u.ac.jp>
Message-ID: <DF855963-4F91-42B3-8754-C12DDC0E1A67@tnw.utwente.nl>

OK, it turns out that if I use a randomly generated matrix I get about  
a 1 minute svd, which corresponds nicely to the value Robert found.

After a bit of sleuthing, I discovered that the zero packing to make  
the matrix have dimensions that were a power of two were the culprit.  
With the packing removed everything suddenly started working again.  
Now I just have to remove all the important bugs (like why I get  
nonphysical results and noncovergence) :)
Thanks for all your help

Cheers
Chris


On Sep 17, 2008, at 10:12 AM, David Cournapeau wrote:

> Chris Lee wrote:
>> Hi All,
>>
>> I have code that needs to repeatedly perform an svd on a matrix  
>> (2304,
>> 2304) shape. It seems to take forever to do (I haven't timed it, but
>> approximately two hours passed on the first trial (without  
>> completion)
>> and I am now running a second to make sure nothing went wrong.
>>
>> Is this because I am not linked to an optimized BLAS/LAPACK library  
>> or
>> something else?
>
> Have you tested your scipy atlas ? 2 hours sounds awfully long, and  
> this
> may be due to some bugs caused during a wrong build (for example,  
> errors
> in the masrhalling of arguments at the C/Fortran interface),
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

***************************************************
Chris Lee
Laser Physics and Nonlinear Optics Group
MESA+ Research Institute for Nanotechnology
University of Twente
Phone: ++31 (0)53 489 3968
fax: ++31 (0)53 489 1102
***************************************************





From gael.varoquaux at normalesup.org  Wed Sep 17 09:57:14 2008
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Wed, 17 Sep 2008 15:57:14 +0200
Subject: [SciPy-user] Error in nonlinear least squares fit analysis
In-Reply-To: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>
References: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>
Message-ID: <20080917135714.GF752@phare.normalesup.org>

On Tue, Sep 16, 2008 at 12:06:32PM -0500, David Lonie wrote:
> a) fmin vs. leastsq:
> The method I wrote ended up using the fmin() function to minimize the
> error vector. What is the difference between fmin and leastsq? Is
> there an advantage to using either?

AFAIK, fmin is a scalar optimizer, where leastsq is a vector optimizer,
using an optimized algorithm to minimize the norm of a vector (
http://en.wikipedia.org/wiki/Levenberg-Marquardt_algorithm ). Leastsq
will thus be more efficient on this problem set.

I am not terribly knowledgeable in this area, so I would appreciate being
corrected if I am talking nonsens.

Ga?l


From bsouthey at gmail.com  Wed Sep 17 11:04:49 2008
From: bsouthey at gmail.com (Bruce Southey)
Date: Wed, 17 Sep 2008 10:04:49 -0500
Subject: [SciPy-user] Error in nonlinear least squares fit analysis
In-Reply-To: <20080917135714.GF752@phare.normalesup.org>
References: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>
	<20080917135714.GF752@phare.normalesup.org>
Message-ID: <48D11C91.8010506@gmail.com>

Gael Varoquaux wrote:
> On Tue, Sep 16, 2008 at 12:06:32PM -0500, David Lonie wrote:
>   
>> a) fmin vs. leastsq:
>> The method I wrote ended up using the fmin() function to minimize the
>> error vector. What is the difference between fmin and leastsq? Is
>> there an advantage to using either?
>>     
>
> AFAIK, fmin is a scalar optimizer, where leastsq is a vector optimizer,
> using an optimized algorithm to minimize the norm of a vector (
> http://en.wikipedia.org/wiki/Levenberg-Marquardt_algorithm ). Leastsq
> will thus be more efficient on this problem set.
>
> I am not terribly knowledgeable in this area, so I would appreciate being
> corrected if I am talking nonsens.
>
> Ga?l
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   

To complete this, fmin uses 'downhill simplex algorithm' (Nelder-Mead 
Simplex algorithm http://en.wikipedia.org/wiki/Nelder-Mead_method).

The big difference is that simplex doesn't use derivatives but 
Levenberg-Marquardt requires first order derivatives. So obviously you 
can not use Levenberg-Marquardt if you don't have the derivatives or 
these are very hard or slow to compute.  Levenberg-Marquardt is likely 
to be faster to converge (but slower because it has to compute 
derivatives) than using simplex or similar methods. However simplex 
methods are more likely than other algorithms to find local maxima 
rather than global maxima so you do need to check for that.

Apart from that, you probably are not missing much.

I have not dealt with non-linear problems in ages to answer the second 
part of your question.  Basically you need the variance of the estimate 
but that very much depends on the type of problem you have.

Bruce



From peridot.faceted at gmail.com  Wed Sep 17 12:00:09 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Wed, 17 Sep 2008 12:00:09 -0400
Subject: [SciPy-user] Gaussian quadrature error
In-Reply-To: <0F8253EA348F49F6A3C3A0AEFE044866@GatewayLaptop>
References: <0F8253EA348F49F6A3C3A0AEFE044866@GatewayLaptop>
Message-ID: <ce557a360809170900r4f2105b6g24089a81a6612150@mail.gmail.com>

2008/9/14 Dan Murphy <chiefmurph at comcast.net>:
> I am trying out the integrate.quadrature function on the function f(x)=e**x
> to the left of the y-axis. If the lower bound in not too negative, I get a
> reasonable answer, but if the lower bound is too negative, I get 0.0 as the
> value of the integral. Here is the code:
>
>
>
> from scipy import *
>
>
>
> def f(x):
>
>     return e**x
>
>
>
> integrate.quadrature(f,-10.0,0.0) # answer is (0.999954600065,
> 3.14148596026e-010)
>
>
>
> but
>
>
>
> integrate.quadrature(f,-1000.0,0.0) # yields (8.35116510531e-090,
> 8.35116510531e-090)
>
>
>
> Note that 'val' and 'err' are equal. Is this a bug in quadrature?

No, unfortunately. It is a limitation of numerical quadrature in
general. Specifically, no matter how adaptive the algorithm is, it can
only base its result on a finite number of sampled points of the
function. If these points are all zero to numerical accuracy, then the
answer must be zero. So if you imagine those samples are taken at the
midpoints of 10 intervals evenly spaced between -1000 and 0, then the
rightmost one returns a value of e**(-50), which is as close to zero
as makes no nevermind. You might be all right if this were an adaptive
scheme and if it used the endpoints, since one endpoint is guaranteed
to give you one. But not using the endpoints is a design feature of
some numerical integration schemes.

The take-home lesson is that you can't just use numerical quadrature
systems blindly; you have to know the features and limitations of the
particular one you're using. Gaussian quadrature can be very accurate
for smooth functions, but it has a very specific domain of
applicability. scipy.optimize.quad is a little more general-purpose by
intent (and necessarily a little less efficient when Gaussian
quadrature will do) but it can be tricked too.

A more specific take-home lesson is to try to normalize your problem
as much as possible, so that all quantities you feed your integrator
are of order unity. Yes, it's a pain to have to handle scale factors
yourself, particularly in the normal case when you're solving a family
of related problems. But you'll get much more reliable performance.

Anne


From rob.clewley at gmail.com  Wed Sep 17 13:42:50 2008
From: rob.clewley at gmail.com (Rob Clewley)
Date: Wed, 17 Sep 2008 13:42:50 -0400
Subject: [SciPy-user] Error in nonlinear least squares fit analysis
In-Reply-To: <48D11C91.8010506@gmail.com>
References: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>
	<20080917135714.GF752@phare.normalesup.org>
	<48D11C91.8010506@gmail.com>
Message-ID: <a749952d0809171042o45a29528r4a59871ce7411e1c@mail.gmail.com>

> I have not dealt with non-linear problems in ages to answer the second
> part of your question.  Basically you need the variance of the estimate
> but that very much depends on the type of problem you have.
>
> Bruce

I'm not a great expert on this theory, so some of my explanation might
get a bit shaky... But, the variance comes from the Hessian, which is
the matrix of second derivatives (i.e., deriv of the Jacobian).

This is nasty to compute using forward differencing (loss of
significance in the numerics) if you don't have an explicit Jacobian,
but a close approximation is usually multiply(transpose(J),J) (this
basically comes down to Taylor's theorem, IIRC). The more locally
quadratic is your residual function then the larger the values will
be, and the variance will be smaller. Small values will mean the
function is flatter locally so you aren't getting as tight a fit.
Different entries in that matrix are basically measuring the curvature
in different cross-sections of the space, I think.

-Rob


From robert.kern at gmail.com  Wed Sep 17 16:08:08 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 17 Sep 2008 15:08:08 -0500
Subject: [SciPy-user] Error in nonlinear least squares fit analysis
In-Reply-To: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>
References: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>
Message-ID: <3d375d730809171308i12f45decv5e6d161c221d88d3@mail.gmail.com>

On Tue, Sep 16, 2008 at 12:06, David Lonie <loniedavid at gmail.com> wrote:
> b) Error in the parameters:
> I'd like to know the precision that the fitted parameters are good to.
> Basically, I'd like to know that b = 3.456 +/- 0.003 instead of just b
> = 3.456.
> leastsq can return a Jacobian matrix -- will pulling out the diagonal
> elements of this matrix give me the results I want? Or is there a
> better way?

leastsq(full_output=True) will also return an estimate of the
covariance matrix in recent (probably > 0.6) SVN versions of scipy.

You might also want to look at scipy.odr.

-- 
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 loniedavid at gmail.com  Wed Sep 17 16:55:49 2008
From: loniedavid at gmail.com (David Lonie)
Date: Wed, 17 Sep 2008 15:55:49 -0500
Subject: [SciPy-user] Error in nonlinear least squares fit analysis
In-Reply-To: <3d375d730809171308i12f45decv5e6d161c221d88d3@mail.gmail.com>
References: <199bcede0809161006xd29ef45lf732790bd1eb7bf3@mail.gmail.com>
	<3d375d730809171308i12f45decv5e6d161c221d88d3@mail.gmail.com>
Message-ID: <199bcede0809171355r37892954n7487612bac17b3f2@mail.gmail.com>

> leastsq(full_output=True) will also return an estimate of the
> covariance matrix in recent (probably > 0.6) SVN versions of scipy.

This sounds like what I'm looking for -- just to make sure I'm looking
at this right, the diagonal terms of the covariance matrix represent
the variance in the parameters? I.e., for the error in a exponential
fit y = coef[0]*e**(coef[1]*x), the error for coef[1] would be the
sqrt of leastsq's cov_x[1][1], and I can return the best value of the
parameter as:

coef[1] +/- sqrt(cov_x[1][1]) ?

This may be a simple question, but I'm having great difficulty finding
a source that explains the meaning of all of these various matricies
:)

Thanks for the help,

Dave


From wagh.utkarsh at gmail.com  Thu Sep 18 03:31:44 2008
From: wagh.utkarsh at gmail.com (utkarsh wagh)
Date: Thu, 18 Sep 2008 13:01:44 +0530
Subject: [SciPy-user] Regarding Genetic algorithm
Message-ID: <7869a7600809180031j282e0221xc21a67b85d540cd5@mail.gmail.com>

Hi,

Can anyone help me out in using the Genetic algorithm (ga) subpackage in
Scipy. If possible can anyone send me the sample codes

thank you,

-- 
Utkarsh Wagh
IIT Delhi
Contact no: +91 9990646707
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080918/4cfa66f3/attachment.html>

From Roger.Fearick at uct.ac.za  Thu Sep 18 04:44:54 2008
From: Roger.Fearick at uct.ac.za (Roger Fearick)
Date: Thu, 18 Sep 2008 10:44:54 +0200
Subject: [SciPy-user]  Error in nonlinear least squares fit analysis
Message-ID: <48D23126.8130.009D.0@uct.ac.za>

Hi all,

I'm an occasional user of scipy but this is one of the things I have
looked at, so I thought I'd better delurk and comment.

> This sounds like what I'm looking for -- just to make sure I'm looking
> at this right, the diagonal terms of the covariance matrix represent
> the variance in the parameters? I.e., for the error in a exponential
> fit y = coef[0]*e**(coef[1]*x), the error for coef[1] would be the
s> qrt of leastsq's cov_x[1][1], and I can return the best value of the
p> arameter as:

> coef[1] +/- sqrt(cov_x[1][1]) ?

> This may be a simple question, but I'm having great difficulty finding
> a source that explains the meaning of all of these various matricies
> :)

It's not a simple question, actually.
The gory details are in something like Numerical Recipes -- in latest (3rd) edition
Sect 15.6 (15.6.5).

There is a scipy example at www.phy.uct.ac.za/courses/python/examples/moreexamples.html
(Non-linear least squares fit) which shows how to process the output of leastsq.

Roger.


--
Roger Fearick
Department of Physics
University of Cape Town






From lfriedri at imtek.de  Thu Sep 18 05:05:53 2008
From: lfriedri at imtek.de (Lars Friedrich)
Date: Thu, 18 Sep 2008 11:05:53 +0200
Subject: [SciPy-user] optimize.fmin_bfgs, vectorial epsilon
Message-ID: <48D219F1.6030408@imtek.de>

Hello,

the docstring of scipy.optimze.fmin_bfgs says that the epsilon parameter 
can be an ndarray:

epsilon : int or ndarray

I assume, that it should be possible to pass different epsilons for all 
the different elements of x.

But, if I do so, this causes an error, since in optimize.approx_fprime 
there are the lines:

ei[k] = epsilon
grad[k] = (f(*((xk+ei,)+args)) - f0)/epsilon

that assume that epsilon is scalar.

Is this intended? Why does the docstring say 'int', shouldn't this be 
'float'?

Besides that: I get 'Warning: Desired error not necessarily achieved due 
to precision loss' on my optimization problem. Could this be due to the 
fact that my parameters differ by several orders of magnitude? Do I have 
to rescale my error function to cope with that? Or should a 
'per-parameter-epsilon' be enough?

Thanks!
Lars


-- 
Dipl.-Ing. Lars Friedrich

Bio- and Nano-Photonics
Department of Microsystems Engineering -- IMTEK
University of Freiburg
Georges-K?hler-Allee 102
D-79110 Freiburg
Germany

phone: +49-761-203-7531
fax:   +49-761-203-7537
room:  01 088
email: lfriedri at imtek.de


From ajvogel at tuks.co.za  Thu Sep 18 08:38:55 2008
From: ajvogel at tuks.co.za (Adolph J. Vogel)
Date: Thu, 18 Sep 2008 14:38:55 +0200
Subject: [SciPy-user]  Stability of scipy svn
In-Reply-To: <48D219F1.6030408@imtek.de>
References: <48D219F1.6030408@imtek.de>
Message-ID: <200809181438.56732.ajvogel@tuks.co.za>

Hi,

I want to use some of the new sparse matrix features in the 
development version of scipy. My question is simple, how stable is 
the current scipy svn? Are there any show stopper bugs?

regards, Adolph



-- 
______________________________________________
Adolph J. Vogel             ajvogel at tuks.co.za
       BEng(Mech)(Pta)            072 592 5836
                                  012 420 4762

    Department Mechanical and Aeronautical 
                Engineering
           University of Pretoria
               South Africa
_______________________________________________


From wnbell at gmail.com  Thu Sep 18 09:57:12 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Thu, 18 Sep 2008 09:57:12 -0400
Subject: [SciPy-user] Stability of scipy svn
In-Reply-To: <200809181438.56732.ajvogel@tuks.co.za>
References: <48D219F1.6030408@imtek.de> <200809181438.56732.ajvogel@tuks.co.za>
Message-ID: <d05265cb0809180657t2712ebe4w96b9022b700edb81@mail.gmail.com>

On Thu, Sep 18, 2008 at 8:38 AM, Adolph J. Vogel <ajvogel at tuks.co.za> wrote:
>
> I want to use some of the new sparse matrix features in the
> development version of scipy. My question is simple, how stable is
> the current scipy svn? Are there any show stopper bugs?
>

I'm not aware of any major bugs in scipy.sparse.  If you find one, let
us know and we'll fix it immediately.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From loniedavid at gmail.com  Thu Sep 18 10:42:25 2008
From: loniedavid at gmail.com (David Lonie)
Date: Thu, 18 Sep 2008 09:42:25 -0500
Subject: [SciPy-user] Error in nonlinear least squares fit analysis
In-Reply-To: <48D23126.8130.009D.0@uct.ac.za>
References: <48D23126.8130.009D.0@uct.ac.za>
Message-ID: <199bcede0809180742i4c48a179o6c7dae3af445a853@mail.gmail.com>

> There is a scipy example at www.phy.uct.ac.za/courses/python/examples/moreexamples.html
> (Non-linear least squares fit) which shows how to process the output of leastsq.

This is exactly what I was looking for! Thanks for the help :)

Dave


From nwagner at iam.uni-stuttgart.de  Thu Sep 18 12:19:25 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 18 Sep 2008 18:19:25 +0200
Subject: [SciPy-user] Regarding Genetic algorithm
In-Reply-To: <7869a7600809180031j282e0221xc21a67b85d540cd5@mail.gmail.com>
References: <7869a7600809180031j282e0221xc21a67b85d540cd5@mail.gmail.com>
Message-ID: <web-111511312@uni-stuttgart.de>

On Thu, 18 Sep 2008 13:01:44 +0530
  "utkarsh wagh" <wagh.utkarsh at gmail.com> wrote:
> Hi,
> 
> Can anyone help me out in using the Genetic algorithm 
>(ga) subpackage in
> Scipy. If possible can anyone send me the sample codes
> 
> thank you,
> 
> -- 
> Utkarsh Wagh
> IIT Delhi
> Contact no: +91 9990646707

See

http://projects.scipy.org/scipy/scipy/ticket/484

Nils
  


From kianatoufighi at hotmail.com  Thu Sep 18 12:53:00 2008
From: kianatoufighi at hotmail.com (Kiana Toufighi)
Date: Thu, 18 Sep 2008 12:53:00 -0400
Subject: [SciPy-user] having problems with fortran compiler on Mac OS X
	10.5.2
Message-ID: <BAY132-W31DAD842550C525D883339A44F0@phx.gbl>


Hi all.  

I am trying to install SciPy on my PowerMac with OS X 10.5.2 (Leopard) by following the instructions here http://www.scipy.org/Installing_SciPy/Mac_OS_X and I'm getting errors with respect to the fortran compiler. 

I  have personally installed the newest version of the fortran compiler v. 4.2.3 and I obviously have a C compiler. 
gcc --version: 
i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5465)
Copyright (C) 2005 Free Software Foundation, Inc.

gfortran --version:
GNU Fortran (GCC) 4.2.3
Copyright (C) 2007 Free Software Foundation, Inc.

Now I've tried building scipy by the following commands and here are the errors I get. 

python setup.py build
Couldn't match compiler version for 'GNU Fortran (GCC) 4.2.3\nCopyright (C) 2007 Free Software Foundation, Inc.

python setup.py config_fc --fcompiler=gfortran build
error: don't know how to compile Fortran code on platform 'posix' with 'gfortran' compiler. Supported compilers are: compaq,none,absoft,intel,f,gnu,sun,nag,vast,ibm,gnu95,intelv,g95,intele,pg,lahey,compaqv,mips,hpux,intelev,intelem)

So then I try one of these listend supported compilers like gnu95 but it's not installed!!
I wish Apple would just include a fortran compiler with their dev tools!

Any help/advice is apprecaited. 

Thanks,
Kiana 

_________________________________________________________________

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080918/b464ccbd/attachment.html>

From torii at gmx.com  Thu Sep 18 14:00:33 2008
From: torii at gmx.com (torii at gmx.com)
Date: Thu, 18 Sep 2008 14:00:33 -0400
Subject: [SciPy-user] Width of the gaussian in stats.kde.gaussian_kde ?
Message-ID: <20080918181527.297260@gmx.com>

Dear scipy users,

I used the kernel-density estimate to make some 2D density plots 
(stats.?kde.?gaussian_kde) and I was very happy with the result. 



But when I do the same exercise over a much larger area, I completely lost 
the details I had with my previous analysis... If I understand correctly, 
this is related to the adaptation of the elementary gaussian to the scale 
of my dataset which now includes large areas with almost no data.



Question: is there a way to control the width of the gaussian in 
stats.kde.gaussian_kde? or should I switch to another technique?



Note: I am both a newbie in python and stats...



Anthony


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080918/5a24b8fa/attachment.html>

From robert.kern at gmail.com  Thu Sep 18 15:18:44 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 18 Sep 2008 14:18:44 -0500
Subject: [SciPy-user] having problems with fortran compiler on Mac OS X
	10.5.2
In-Reply-To: <BAY132-W31DAD842550C525D883339A44F0@phx.gbl>
References: <BAY132-W31DAD842550C525D883339A44F0@phx.gbl>
Message-ID: <3d375d730809181218sf81bd04i5d34803b571f8aa8@mail.gmail.com>

On Thu, Sep 18, 2008 at 11:53, Kiana Toufighi <kianatoufighi at hotmail.com> wrote:
> Hi all.
>
> I am trying to install SciPy on my PowerMac with OS X 10.5.2 (Leopard) by
> following the instructions here
> http://www.scipy.org/Installing_SciPy/Mac_OS_X and I'm getting errors with
> respect to the fortran compiler.
>
> I  have personally installed the newest version of the fortran compiler v.
> 4.2.3 and I obviously have a C compiler.
> gcc --version:
> i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5465)
> Copyright (C) 2005 Free Software Foundation, Inc.
>
> gfortran --version:
> GNU Fortran (GCC) 4.2.3
> Copyright (C) 2007 Free Software Foundation, Inc.
>
> Now I've tried building scipy by the following commands and here are the
> errors I get.
>
> python setup.py build
> Couldn't match compiler version for 'GNU Fortran (GCC) 4.2.3\nCopyright (C)
> 2007 Free Software Foundation, Inc.
>
> python setup.py config_fc --fcompiler=gfortran build
> error: don't know how to compile Fortran code on platform 'posix' with
> 'gfortran' compiler. Supported compilers are:
> compaq,none,absoft,intel,f,gnu,sun,nag,vast,ibm,gnu95,intelv,g95,intele,pg,lahey,compaqv,mips,hpux,intelev,intelem)
>
> So then I try one of these listend supported compilers like gnu95 but it's
> not installed!!
> I wish Apple would just include a fortran compiler with their dev tools!

--fcompiler=gnu95 actually refers to gfortran. Can you show us the
output you get when you try to use that?

-- 
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 kianatoufighi at hotmail.com  Thu Sep 18 15:23:44 2008
From: kianatoufighi at hotmail.com (Kiana Toufighi)
Date: Thu, 18 Sep 2008 15:23:44 -0400
Subject: [SciPy-user] having problems with fortran compiler on Mac OS
	X	10.5.2
In-Reply-To: <3d375d730809181218sf81bd04i5d34803b571f8aa8@mail.gmail.com>
References: <BAY132-W31DAD842550C525D883339A44F0@phx.gbl>
	<3d375d730809181218sf81bd04i5d34803b571f8aa8@mail.gmail.com>
Message-ID: <BAY132-W4695CC5EAAE23416B4ED4EA44F0@phx.gbl>


Hi Robert,

Thanks for the reply. When I try building with python setup.py config_fc --fcompiler=gnu95 build
the last few lines of output are:

running build_clib
customize UnixCCompiler
customize UnixCCompiler using build_clib
Could not locate executable g77
Could not locate executable f77
Could not locate executable f95
customize Gnu95FCompiler
Couldn't match compiler version for 'GNU Fortran (GCC) 4.2.3\nCopyright (C) 2007 Free Software Foundation, Inc.\n\nGNU Fortran comes with NO WARRANTY, to the extent permitted by law.\nYou may redistribute copies of GNU Fortran\nunder the terms of the GNU General Public License.\nFor more information about these matters, see the file named COPYING\n'
customize Gnu95FCompiler using build_clib
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
Couldn't match compiler version for 'GNU Fortran (GCC) 4.2.3\nCopyright (C) 2007 Free Software Foundation, Inc.\n\nGNU Fortran comes with NO WARRANTY, to the extent permitted by law.\nYou may redistribute copies of GNU Fortran\nunder the terms of the GNU General Public License.\nFor more information about these matters, see the file named COPYING\n'
warning: build_ext: fcompiler=gnu95 is not available.


Thanks,
K.

_________________________________________________________________

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080918/6b08d96e/attachment.html>

From peridot.faceted at gmail.com  Thu Sep 18 15:41:40 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Thu, 18 Sep 2008 15:41:40 -0400
Subject: [SciPy-user] Width of the gaussian in stats.kde.gaussian_kde ?
In-Reply-To: <20080918181527.297260@gmx.com>
References: <20080918181527.297260@gmx.com>
Message-ID: <ce557a360809181241l3d21e806lae21c8af94182165@mail.gmail.com>

2008/9/18  <torii at gmx.com>:

> I used the kernel-density estimate to make some 2D density plots
> (stats.kde.gaussian_kde) and I was very happy with the result.
>
> But when I do the same exercise over a much larger area, I completely lost
> the details I had with my previous analysis... If I understand correctly,
> this is related to the adaptation of the elementary gaussian to the scale of
> my dataset which now includes large areas with almost no data.
>
> Question: is there a way to control the width of the gaussian in
> stats.kde.gaussian_kde? or should I switch to another technique?
>
> Note: I am both a newbie in python and stats...

Kernel-density estimates approximate the distribution of your data by
placing a copy of the kernel at each data point. Scipy's gaussian_kde
uses multidimensional gaussians as the kernel. It provides various
features, including reasonably efficient evaluation, integration over
boxes and against gaussians and other gaussian KDEs, and most
relevantly, automatic selection of the covariance matrix of the
kernel. There are a number of different ways to do this automatic
selection in the statistical literature, and the one implemented in
gaussian_kde is appropriate for a unimodal distribution: it uses the
covariance matrix of your data, scaled by a factor depending on the
dimensionality and number of data points. If, however, you have a data
set consisting of several narrow widely-separated peaks, this will
give a needlessly blurry estimate.

A more powerful standard tool for choosing the variance in one
dimension is to use a "plug-in estimator". The idea is to try to
choose the variance that minimizes the estimated mean-squared error.
The mean squared error comes from two things: the randomness of the
original sampling, and the smoothing by the gaussians. If you choose a
very broad kernel, you smooth out almost all the noise, but you have
lots of mean-squared error because you've smoothed away any features
the distribution really had. On the other hand, if you use a very
narrow kernel, you haven't smoothed the distribution much at all, but
now the randomness of your points wrecks things. So you have to select
some optimal kernel. Ideally, you could run a numerical minimization
by evaluating the mean squared error for each trial variance.
Unfortunately this requires you to know the true distribution. That's
where the "plug-in estimator" comes in: you choose some crude way to
estimate the distribution, and use this crude estimate in your
mean-squared error calculations to get the best variance. You then use
this best variance as your kernel width.

What does this have to do with you? Well, unfortunately, plug-in
estimators are not implemented in scipy, probably in part because it's
difficult enough in one dimension, and a real horror in several. You
could try to implement it (after doing some reading in the statistical
literature) but I suspect that's not how you want to address the
problem.

I suggest you choose a "representative" part of your data that
consists of a single peak, feed it into gaussian_kde, and extract the
variance. Then create a gaussian_kde for the whole data set using that
same variance. You can optionally fudge the covariance as necessary in
between.

The first part, extracting the covariance from a gaussian_kde
instance, should be easy: as soon as you've constructed the
gaussian_kde it is stored in the covariance attribute. The second
part, building a gaussian_kde instance with set covariance, is going
to require a little hacking. Here's how I'd do it (untested):

class gaussian_kde_set_covariance(scipy.stats.gaussian_kde):
    def __init__(self, dataset, covariance):
        self.covariance = covariance
        scipy.stats.gaussian_kde.__init__(self, dataset)

    def _compute_covariance(self):
        self.inv_cov = linalg.inv(self.covariance)
        self._norm_factor = sqrt(linalg.det(2*pi*self.covariance)) * self.n

This creates a derived class from gaussian_kde which does everything
the same way except for how it is constructed and how it computes its
covariance matrix (i.e. it doesn't, but it does compute the various
other matrices it needs).

Good luck,
Anne


From robert.kern at gmail.com  Thu Sep 18 16:11:22 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 18 Sep 2008 15:11:22 -0500
Subject: [SciPy-user] having problems with fortran compiler on Mac OS X
	10.5.2
In-Reply-To: <BAY132-W4695CC5EAAE23416B4ED4EA44F0@phx.gbl>
References: <BAY132-W31DAD842550C525D883339A44F0@phx.gbl>
	<3d375d730809181218sf81bd04i5d34803b571f8aa8@mail.gmail.com>
	<BAY132-W4695CC5EAAE23416B4ED4EA44F0@phx.gbl>
Message-ID: <3d375d730809181311n543469ecs4b4b8ef41d295f7a@mail.gmail.com>

On Thu, Sep 18, 2008 at 14:23, Kiana Toufighi <kianatoufighi at hotmail.com> wrote:
> Hi Robert,
>
> Thanks for the reply. When I try building with python setup.py config_fc
> --fcompiler=gnu95 build
> the last few lines of output are:
>
> running build_clib
> customize UnixCCompiler
> customize UnixCCompiler using build_clib
> Could not locate executable g77
> Could not locate executable f77
> Could not locate executable f95
> customize Gnu95FCompiler
> Couldn't match compiler version for 'GNU Fortran (GCC) 4.2.3\nCopyright (C)
> 2007 Free Software Foundation, Inc.\n\nGNU Fortran comes with NO WARRANTY,
> to the extent permitted by law.\nYou may redistribute copies of GNU
> Fortran\nunder the terms of the GNU General Public License.\nFor more
> information about these matters, see the file named COPYING\n'

What version of numpy do you have? I use this version of gfortran on
10.5.2, too, with SVN numpy, but it looks like numpy 1.1.1 has the
same version-processing code. numpy 1.2.0 will be out very shortly,
too.

-- 
Robert Kern

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


From dwf at cs.toronto.edu  Thu Sep 18 17:23:50 2008
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 18 Sep 2008 17:23:50 -0400
Subject: [SciPy-user] scipy build fails on Linux-x86-64 with custom ATLAS
Message-ID: <D68714FB-1A37-4F6F-B49D-BD6381BFFE8F@cs.toronto.edu>

Hi folks,

  I am having an issue building scipy from SVN. I rolled my own ATLAS  
and installed it under /usr/local/atlas, and passed the ATLAS  
environment variable. I also told it to use gfortran as opposed to  
g77 as that's what was used for ATLAS.

The compilation error is as follows:

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

/usr/bin/gfortran -Wall -Wall -shared build/temp.linux-x86_64-2.5/ 
scipy/integrate/_odepackmodule.o -L/usr/local/atlas/lib -Lbuild/ 
temp.linux-x86_64-2.5 -lodepack -llinpack_lite -lmach -lptf77blas - 
lptcblas -latlas -lgfortran -o build/lib.linux-x86_64-2.5/scipy/ 
integrate/_odepack.so
/usr/bin/ld: /usr/local/atlas/lib/libptf77blas.a(dscal.o): relocation  
R_X86_64_32 against `a local symbol' can not be used when making a  
shared object; recompile with -fPIC
/usr/local/atlas/lib/libptf77blas.a: could not read symbols: Bad value
collect2: ld returned 1 exit status
/usr/bin/ld: /usr/local/atlas/lib/libptf77blas.a(dscal.o): relocation  
R_X86_64_32 against `a local symbol' can not be used when making a  
shared object; recompile with -fPIC
/usr/local/atlas/lib/libptf77blas.a: could not read symbols: Bad value
collect2: ld returned 1 exit status
error: Command "/usr/bin/gfortran -Wall -Wall -shared build/ 
temp.linux-x86_64-2.5/scipy/integrate/_odepackmodule.o -L/usr/local/ 
atlas/lib -Lbuild/temp.linux-x86_64-2.5 -lodepack -llinpack_lite - 
lmach -lptf77blas -lptcblas -latlas -lgfortran -o build/lib.linux- 
x86_64-2.5/scipy/integrate/_odepack.so" failed with exit status 1

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

I notice it's complaining about -fPIC, except that when I configured  
ATLAS, I specifically told it that every compiler should use -fPIC:

	../configure -m 2400 -b 64 -D c -DPentiumCPS=2400 --prefix=/usr/ 
local/atlas --with-netlib-lapack=../../lapack-3.1.1/lapack_LINUX.a - 
Fa acg '-fPIC'

That last option, if I am reading the docs correctly, should take  
care of it, yet SciPy's build still fails. Does anyone have any idea  
why?

Thanks,

David


From robince at gmail.com  Thu Sep 18 17:31:19 2008
From: robince at gmail.com (Robin)
Date: Thu, 18 Sep 2008 22:31:19 +0100
Subject: [SciPy-user] scipy build fails on Linux-x86-64 with custom ATLAS
In-Reply-To: <D68714FB-1A37-4F6F-B49D-BD6381BFFE8F@cs.toronto.edu>
References: <D68714FB-1A37-4F6F-B49D-BD6381BFFE8F@cs.toronto.edu>
Message-ID: <b14b627d0809181431u1669b100k7c63154bd82c8dec@mail.gmail.com>

On Thu, Sep 18, 2008 at 10:23 PM, David Warde-Farley <dwf at cs.toronto.edu> wrote:
>        ../configure -m 2400 -b 64 -D c -DPentiumCPS=2400 --prefix=/usr/
> local/atlas --with-netlib-lapack=../../lapack-3.1.1/lapack_LINUX.a -
> Fa acg '-fPIC'

I've always used -Fa alg '-fPIC' - this is also suggested on the wiki page:
http://scipy.org/Installing_SciPy/Linux

Perhaps this could make a difference? I think acg is only C compilers,
while the l includes fortran as well - since the error your getting is
coming from a Fortran library I think it's likely this could fix it.

Cheers

Robin


From bblais at bryant.edu  Thu Sep 18 17:33:35 2008
From: bblais at bryant.edu (Brian Blais)
Date: Thu, 18 Sep 2008 17:33:35 -0400
Subject: [SciPy-user] delays and odeint
Message-ID: <F0CD4867-2062-491B-AC70-C8A943AE535A@bryant.edu>

Hello,

I am trying to do some dynamics systems modeling, specifically  
modeling population dynamics.  One piece of the model involves a  
delay, where you have something like:

dy=x-delay(x,10)

where a population, y, has an amount x added to it and an amount of  
x, from 10 time steps before, exiting.  In this way the population,  
y, had members adding to it that last in population y for 10 time  
steps and then exit.  Is there a good way of representing this in  
such a way that I can still use odeint to solve the resulting diff- 
eqs?  I could write an Euler-method myself, and represent the delay  
as a list which I push from one end and pop from the other, but that  
seems a little hacky.  odeint uses an adaptive step-size, so that  
solution wouldn't work.  I wasn't sure if there already existed a  
tool for helping with this sort of thing.


thanks,

		Brian Blais

-- 
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/20080918/2462965f/attachment.html>

From dwf at cs.toronto.edu  Thu Sep 18 17:56:05 2008
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 18 Sep 2008 17:56:05 -0400
Subject: [SciPy-user] scipy build fails on Linux-x86-64 with custom ATLAS
In-Reply-To: <b14b627d0809181431u1669b100k7c63154bd82c8dec@mail.gmail.com>
References: <D68714FB-1A37-4F6F-B49D-BD6381BFFE8F@cs.toronto.edu>
	<b14b627d0809181431u1669b100k7c63154bd82c8dec@mail.gmail.com>
Message-ID: <FB0D58B3-E1BB-4972-91A8-08D58E50CB14@cs.toronto.edu>


On 18-Sep-08, at 5:31 PM, Robin wrote:

> On Thu, Sep 18, 2008 at 10:23 PM, David Warde-Farley  
> <dwf at cs.toronto.edu> wrote:
>>        ../configure -m 2400 -b 64 -D c -DPentiumCPS=2400 --prefix=/ 
>> usr/
>> local/atlas --with-netlib-lapack=../../lapack-3.1.1/lapack_LINUX.a -
>> Fa acg '-fPIC'
>
> I've always used -Fa alg '-fPIC' - this is also suggested on the  
> wiki page:
> http://scipy.org/Installing_SciPy/Linux
>
> Perhaps this could make a difference? I think acg is only C compilers,
> while the l includes fortran as well - since the error your getting is
> coming from a Fortran library I think it's likely this could fix it.

I just figured that out myself. As usual, one character makes all the  
difference. :(

That, and I didn't know such a wiki page existed. Oh well. Problem  
solved.

Thanks for the quick reply,

David


From gdahl at cs.toronto.edu  Thu Sep 18 20:26:01 2008
From: gdahl at cs.toronto.edu (gdahl at cs.toronto.edu)
Date: Thu, 18 Sep 2008 20:26:01 -0400 (EDT)
Subject: [SciPy-user] can't build scipy on mac os 10.5.5 intel
Message-ID: <27496.67.212.30.212.1221783961.squirrel@webmail.cs.toronto.edu>

Hi everyone.
I am not sure if my other email account is working so I apologize if this
already hit the list.
I get the error below when I run python setup.py
build with the latest SVN version of scipy.
I have numpy (latest svn) and it works in python2.5, which is the
version of python
I want to use scipy from.  I have gcc 4.0.1 and gfortran 4.2.1.

This is the command I ran:
sudo python setup.py build_src build_clib --fcompiler=gnu95 build_ext
--fcompiler=gnu95 build

I have pasted some of the output of python setup.py build below.  I don't
really know what I am doing, but I would really like to get
scipy working and I would appreciate any help anyone can give me!  I
looked around in the archives with google a bit, but nothing seemed to
deal with my situation exactly, hopefully I didn't miss anything.
- George


$ sudo python setup.py build_src build_clib --fcompiler=gnu95
build_ext --fcompiler=gnu95 build
.
.
.
building extension "scipy.sparse.linalg.dsolve.umfpack.__umfpack" sources
  adding 'scipy/sparse/linalg/dsolve/umfpack/umfpack.i' to sources.
swig: scipy/sparse/linalg/dsolve/umfpack/umfpack.i
swig -python -o
build/src.macosx-10.3-i386-2.5/scipy/sparse/linalg/dsolve/umfpack/_umfpack_wrap.c
-outdir build/src.macosx-10.3-i386-2.5/scipy/sparse/linalg/dsolve/umfpack
scipy/sparse/linalg/dsolve/umfpack/umfpack.i
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to find
'umfpack.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:193: Error: Unable to find
'umfpack_solve.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:194: Error: Unable to find
'umfpack_defaults.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:195: Error: Unable to find
'umfpack_triplet_to_col.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:196: Error: Unable to find
'umfpack_col_to_triplet.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:197: Error: Unable to find
'umfpack_transpose.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:198: Error: Unable to find
'umfpack_scale.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:200: Error: Unable to find
'umfpack_report_symbolic.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:201: Error: Unable to find
'umfpack_report_numeric.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:202: Error: Unable to find
'umfpack_report_info.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:203: Error: Unable to find
'umfpack_report_control.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:215: Error: Unable to find
'umfpack_symbolic.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:216: Error: Unable to find
'umfpack_numeric.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:225: Error: Unable to find
'umfpack_free_symbolic.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:226: Error: Unable to find
'umfpack_free_numeric.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:248: Error: Unable to find
'umfpack_get_lunz.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:272: Error: Unable to find
'umfpack_get_numeric.h'
error: command 'swig' failed with exit status 1





From torii at gmx.com  Thu Sep 18 22:44:09 2008
From: torii at gmx.com (Anthony)
Date: Fri, 19 Sep 2008 02:44:09 +0000 (UTC)
Subject: [SciPy-user]
	=?utf-8?q?Width_of_the_gaussian_in_stats=2Ekde=2Egau?=
	=?utf-8?q?ssian=5Fkde_=3F?=
References: <20080918181527.297260@gmx.com>
	<ce557a360809181241l3d21e806lae21c8af94182165@mail.gmail.com>
Message-ID: <loom.20080919T023113-707@post.gmane.org>

Thank you so much for taking the time to explain all this things. 
Your suggestion makes sense but I am not sure I can implement it... I'll try it
though.
Cheers,
Anthony




From rob.clewley at gmail.com  Fri Sep 19 00:32:52 2008
From: rob.clewley at gmail.com (Rob Clewley)
Date: Fri, 19 Sep 2008 00:32:52 -0400
Subject: [SciPy-user] delays and odeint
In-Reply-To: <F0CD4867-2062-491B-AC70-C8A943AE535A@bryant.edu>
References: <F0CD4867-2062-491B-AC70-C8A943AE535A@bryant.edu>
Message-ID: <a749952d0809182132k67cbde50jb179f8b701902552@mail.gmail.com>

> I am trying to do some dynamics systems modeling, specifically modeling
> population dynamics.  One piece of the model involves a delay, where you
> have something like:
> dy=x-delay(x,10)

If you just want to solve the problem, I actually recommend using the
XPPAUT program. It's great for that kind of thing, is easy to install,
and even has a graphical interface. Unfortunately I can't say my own
python dynamics package has support for delayed systems yet, although
I think PyDDE looks good if you must have a python solution.

-Rob


From mnandris at blueyonder.co.uk  Fri Sep 19 05:58:50 2008
From: mnandris at blueyonder.co.uk (Michael)
Date: Fri, 19 Sep 2008 10:58:50 +0100
Subject: [SciPy-user] How to get a ppf for scipy.stats.beta
Message-ID: <1221818330.6174.2.camel@mik>

Hi list,

i am trying to use the percent point function to sample from the cdf for
a beta distribution

beta.ppf() returns an array instead of x; also you cannot create a ppf
using the same idiom as beta.cdf(x,a,b,size=n)

How do i go about getting the ppf for a beta dist? 

thanks in advance - much hairpulling on this <grr>

Michael


from scipy.stats import norm
from scipy.stats import beta
from scipy import linspace

print norm.cdf(1.2)		# 0.884930329778 
print norm.ppf(0.884930329778)	# 1.2

a=2
b=7 
n=10e3
x=linspace(0,1,n)

cdf=beta.cdf(x,a,b,size=n)
print cdf[2371]			# 0.6

B=beta(x,a,b,size=n)

print B.cdf(2371)	
print B.ppf(0.6)		# <-- should be 2371?
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080919/853ecae6/attachment.sig>

From massimo.sandal at unibo.it  Fri Sep 19 08:00:02 2008
From: massimo.sandal at unibo.it (massimo sandal)
Date: Fri, 19 Sep 2008 14:00:02 +0200
Subject: [SciPy-user] Regarding Genetic algorithm
In-Reply-To: <web-111511312@uni-stuttgart.de>
References: <7869a7600809180031j282e0221xc21a67b85d540cd5@mail.gmail.com>
	<web-111511312@uni-stuttgart.de>
Message-ID: <48D39442.4080905@unibo.it>

Nils Wagner wrote:
> On Thu, 18 Sep 2008 13:01:44 +0530
>   "utkarsh wagh" <wagh.utkarsh at gmail.com> wrote:
>> Hi,
>>
>> Can anyone help me out in using the Genetic algorithm 
>> (ga) subpackage in
>> Scipy. If possible can anyone send me the sample codes
>>
>> thank you,
>>
>> -- 
>> Utkarsh Wagh
>> IIT Delhi
>> Contact no: +91 9990646707
> 
> See
> 
> http://projects.scipy.org/scipy/scipy/ticket/484

Sorry but I fail to understand how this page can help.

m.

-- 
Massimo Sandal , Ph.D.
University of Bologna
Department of Biochemistry "G.Moruzzi"

snail mail:
Via Irnerio 48, 40126 Bologna, Italy

email:
massimo.sandal at unibo.it

web:
http://www.biocfarm.unibo.it/samori/people/sandal.html

tel: +39-051-2094388
fax: +39-051-2094387
-------------- next part --------------
A non-text attachment was scrubbed...
Name: massimo_sandal.vcf
Type: text/x-vcard
Size: 274 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080919/e1c50903/attachment.vcf>

From david.huard at gmail.com  Fri Sep 19 08:52:31 2008
From: david.huard at gmail.com (David Huard)
Date: Fri, 19 Sep 2008 08:52:31 -0400
Subject: [SciPy-user] How to get a ppf for scipy.stats.beta
In-Reply-To: <1221818330.6174.2.camel@mik>
References: <1221818330.6174.2.camel@mik>
Message-ID: <91cf711d0809190552t2f6f530es3273fe5d2e805f7a@mail.gmail.com>

Hi Michael,

calling beta(a, b) will generate a "frozen distribution", that is one whose
parameters are fixed by you:

B = beta(a, b)   # note that you should't pass x at this stage.

B.pdf(x)
B.cdf(x)

Now if you want to sample from the Beta, simply use the rvs method:

B.rvs(1000)


You can also simply do

beta.rvs(a,b,size=1000)


HTH,

David


On Fri, Sep 19, 2008 at 5:58 AM, Michael <mnandris at blueyonder.co.uk> wrote:

> Hi list,
>
> i am trying to use the percent point function to sample from the cdf for
> a beta distribution
>
> beta.ppf() returns an array instead of x; also you cannot create a ppf
> using the same idiom as beta.cdf(x,a,b,size=n)
>
> How do i go about getting the ppf for a beta dist?
>
> thanks in advance - much hairpulling on this <grr>
>
> Michael
>
>
> from scipy.stats import norm
> from scipy.stats import beta
> from scipy import linspace
>
> print norm.cdf(1.2)             # 0.884930329778
> print norm.ppf(0.884930329778)  # 1.2
>
> a=2
> b=7
> n=10e3
> x=linspace(0,1,n)
>
> cdf=beta.cdf(x,a,b,size=n)
> print cdf[2371]                 # 0.6
>
> B=beta(x,a,b,size=n)
>
> print B.cdf(2371)
> print B.ppf(0.6)                # <-- should be 2371?
>
> _______________________________________________
> 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/20080919/0a02e848/attachment.html>

From josef.pktd at gmail.com  Fri Sep 19 09:40:31 2008
From: josef.pktd at gmail.com (joep)
Date: Fri, 19 Sep 2008 06:40:31 -0700 (PDT)
Subject: [SciPy-user] How to get a ppf for scipy.stats.beta
In-Reply-To: <91cf711d0809190552t2f6f530es3273fe5d2e805f7a@mail.gmail.com>
References: <1221818330.6174.2.camel@mik>
	<91cf711d0809190552t2f6f530es3273fe5d2e805f7a@mail.gmail.com>
Message-ID: <7925319f-150c-42c2-81e0-83576892e404@w7g2000hsa.googlegroups.com>

Scaling is done through the ``scale`` keyword, not through ``size``,
``size`` refers to sample size for rvs

I think what you want is this:

Use distribution function directly, not frozen:

>>> beta.cdf(2371,a,b,scale=n)
array(0.59995766129911532)
>>> beta.ppf(0.6,a,b,scale=n)
array(2371.1617428411146)

Freeze the distribution at the distribution parameters, but not at the
value (x) at which you want to evaluate cdf, ppf,..
In your case the x was fixed at the initially defined values

>>> beta(a,b,scale=n).cdf(2371)
array(0.59995766129911532)
>>> beta(a,b,scale=n).ppf(0.6)
array(2371.1617428411146)

>>> B = beta(a,b,scale=n)
>>> B.cdf(2371)
array(0.59995766129911532)
>>> B.ppf(0.6)
array(2371.1617428411146)

Josef
On Sep 19, 8:52?am, "David Huard" <david.hu... at gmail.com> wrote:
> Hi Michael,
>
> calling beta(a, b) will generate a "frozen distribution", that is one whose
> parameters are fixed by you:
>
> B = beta(a, b) ? # note that you should't pass x at this stage.
>
> B.pdf(x)
> B.cdf(x)
>
> Now if you want to sample from the Beta, simply use the rvs method:
>
> B.rvs(1000)
>
> You can also simply do
>
> beta.rvs(a,b,size=1000)
>
> HTH,
>
> David
>
> On Fri, Sep 19, 2008 at 5:58 AM, Michael <mnand... at blueyonder.co.uk> wrote:
> > Hi list,
>
> > i am trying to use the percent point function to sample from the cdf for
> > a beta distribution
>
> > beta.ppf() returns an array instead of x; also you cannot create a ppf
> > using the same idiom as beta.cdf(x,a,b,size=n)
>
> > How do i go about getting the ppf for a beta dist?
>
> > thanks in advance - much hairpulling on this <grr>
>
> > Michael
>
> > from scipy.stats import norm
> > from scipy.stats import beta
> > from scipy import linspace
>
> > print norm.cdf(1.2) ? ? ? ? ? ? # 0.884930329778
> > print norm.ppf(0.884930329778) ?# 1.2
>
> > a=2
> > b=7
> > n=10e3
> > x=linspace(0,1,n)
>
> > cdf=beta.cdf(x,a,b,size=n)
> > print cdf[2371] ? ? ? ? ? ? ? ? # 0.6
>
> > B=beta(x,a,b,size=n)
>
> > print B.cdf(2371)
> > print B.ppf(0.6) ? ? ? ? ? ? ? ?# <-- should be 2371?
>
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-u... at scipy.org
> >http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-u... at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user


From lroubeyrie at limair.asso.fr  Fri Sep 19 04:14:49 2008
From: lroubeyrie at limair.asso.fr (Lionel Roubeyrie)
Date: Fri, 19 Sep 2008 10:14:49 +0200
Subject: [SciPy-user] Timeseries Cannot specify output type twice
Message-ID: <48D35F79.6050106@limair.asso.fr>

Hi all,
trying to use the last SVN version of the TimeSeries module, I run 
directly into problem on a simple test like this one :

####################################
 >> import numpy as np
 >>> import scikits.timeseries as ts
 >>> data = np.random.uniform(-100,100,600)
 >>> today = ts.now('B')
 >>> series = ts.time_series(data, dtype=np.float_, freq='B', 
start_date=today-600)
 >>> series
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
   File 
"/usr/lib/python2.5/site-packages/scikits/timeseries/tseries.py", line 
715, in __repr__
     return desc_short % {'data': str(self._series),
   File 
"/usr/lib/python2.5/site-packages/scikits/timeseries/tseries.py", line 
485, in _get_series
     return self.view(MaskedArray)
   File 
"/usr/lib/python2.5/site-packages/scikits/timeseries/tseries.py", line 
471, in view
     output = MaskedArray.view(self, dtype=dtype, type=type)
ValueError: Cannot specify output type twice.
 >>> np.__version__
'1.1.1'
 >>> ts.__version__
'0.67.0.dev-r1404'

####################################
Is it a know problem or this problem comes from my installation?
Thanks for your help

-- 
Lionel Roubeyrie - lroubeyrie at limair.asso.fr
Charg? d'?tudes et de maintenance
LIMAIR - la Surveillance de l'Air en Limousin
http://www.limair.asso.fr




From pgmdevlist at gmail.com  Fri Sep 19 12:18:20 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Fri, 19 Sep 2008 12:18:20 -0400
Subject: [SciPy-user] Timeseries Cannot specify output type twice
In-Reply-To: <48D35F79.6050106@limair.asso.fr>
References: <48D35F79.6050106@limair.asso.fr>
Message-ID: <200809191218.21054.pgmdevlist@gmail.com>

On Friday 19 September 2008 04:14:49 Lionel Roubeyrie wrote:
> Hi all,
> trying to use the last SVN version of the TimeSeries module, I run
> directly into problem on a simple test like this one :

That's a known problem, sorry about that. You'll need numpy 1.2.1 (I know, 
1.2.0 iss not even released yet...), or the latest SVN.
Alternatively, just replace line 471 of tseries.py with
>>>?output = MaskedArray.view(self, dtype)

I'll commit a workaround today or tomorrow.
Thanks for your patience.
P.



From nwagner at iam.uni-stuttgart.de  Fri Sep 19 14:01:41 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 19 Sep 2008 20:01:41 +0200
Subject: [SciPy-user] scikits.ann
Message-ID: <web-111572639@uni-stuttgart.de>

Hi all,

I cannot find a component "ann" in the trac
http://scipy.org/scipy/scikits/

Where can I report a bug ?


python setup.py install
running install
Checking .pth file support in 
/usr/local/lib64/python2.5/site-packages
/usr/bin/python -E -c pass
TEST PASSED: /usr/local/lib64/python2.5/site-packages 
appears to support .pth files
running bdist_egg
running egg_info
running build_src
building extension "scikits.ann._ANN" sources
creating build
creating build/src.linux-x86_64-2.5
creating build/src.linux-x86_64-2.5/scikits
creating build/src.linux-x86_64-2.5/scikits/ann
swig++: scikits/ann/ANN.i
swig -python -c++ -Iscikits/ann -I/usr/local/include -o 
build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.cpp 
-outdir build/src.linux-x86_64-2.5/scikits/ann 
scikits/ann/ANN.i
building data_files sources
writing requirements to scikits.ann.egg-info/requires.txt
writing scikits.ann.egg-info/PKG-INFO
writing namespace_packages to 
scikits.ann.egg-info/namespace_packages.txt
writing top-level names to 
scikits.ann.egg-info/top_level.txt
writing dependency_links to 
scikits.ann.egg-info/dependency_links.txt
writing manifest file 'scikits.ann.egg-info/SOURCES.txt'
installing library code to build/bdist.linux-x86_64/egg
running install_lib
running build_py
creating build/lib.linux-x86_64-2.5
creating build/lib.linux-x86_64-2.5/scikits
copying ./scikits/__init__.py -> 
build/lib.linux-x86_64-2.5/scikits
creating build/lib.linux-x86_64-2.5/scikits/ann
copying scikits/ann/info.py -> 
build/lib.linux-x86_64-2.5/scikits/ann
copying scikits/ann/__init__.py -> 
build/lib.linux-x86_64-2.5/scikits/ann
copying scikits/ann/setup.py -> 
build/lib.linux-x86_64-2.5/scikits/ann
copying scikits/ann/version.py -> 
build/lib.linux-x86_64-2.5/scikits/ann
copying build/src.linux-x86_64-2.5/scikits/ann/ANN.py -> 
build/lib.linux-x86_64-2.5/scikits/ann
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
building 'scikits.ann._ANN' extension
compiling C++ sources
C compiler: g++ -pthread -fno-strict-aliasing -DNDEBUG -O2 
-fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -g -fPIC

creating build/temp.linux-x86_64-2.5
creating build/temp.linux-x86_64-2.5/build
creating 
build/temp.linux-x86_64-2.5/build/src.linux-x86_64-2.5
creating 
build/temp.linux-x86_64-2.5/build/src.linux-x86_64-2.5/scikits
creating 
build/temp.linux-x86_64-2.5/build/src.linux-x86_64-2.5/scikits/ann
compile options: '-Iscikits/ann -I/usr/local/include 
-I/usr/local/lib64/python2.5/site-packages/numpy/core/include 
-I/usr/include/python2.5 -c'
g++: build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.cpp
build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.cpp:2563:25: 
error: ANN/ANN.h: No such file or directory
In file included from 
build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.cpp:2564:
scikits/ann/kdtree.h:22:21: error: ANN/ANN.h: No such file 
or directory
scikits/ann/kdtree.h:101:3: warning: no newline at end of 
file
scikits/ann/kdtree.h:27: error: ?ANNpointArray? does not 
name a type
scikits/ann/kdtree.h:28: error: ISO C++ forbids 
declaration of ?ANNkd_tree? with no type
scikits/ann/kdtree.h:28: error: expected ?;? before ?*? 
token
scikits/ann/kdtree.h:43: error: ?ANN_KD_SUGGEST? was not 
declared in this scope
scikits/ann/kdtree.h: In constructor 
?_kdtree::_kdtree(double*, int, int, int, int)?:
scikits/ann/kdtree.h:44: error: ?pts? was not declared in 
this scope
scikits/ann/kdtree.h:44: error: ?annAllocPts? was not 
declared in this scope
scikits/ann/kdtree.h:53: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h:53: error: expected type-specifier 
before ?ANNkd_tree?
scikits/ann/kdtree.h:53: error: expected `;' before 
?ANNkd_tree?
scikits/ann/kdtree.h: In destructor ?_kdtree::~_kdtree()?:
scikits/ann/kdtree.h:57: error: ?pts? was not declared in 
this scope
scikits/ann/kdtree.h:57: error: ?annDeallocPts? was not 
declared in this scope
scikits/ann/kdtree.h:58: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h: In member function ?void 
_kdtree::_knn2(double*, int, int, int, int, int*, int, 
int, double*, double) const?:
scikits/ann/kdtree.h:71: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h: In member function ?const char* 
_kdtree::stringRep(bool) const?:
scikits/ann/kdtree.h:98: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h:98: error: ?ANNtrue? was not declared 
in this scope
scikits/ann/kdtree.h:98: error: ?ANNfalse? was not 
declared in this scope
build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.cpp:2563:25: 
error: ANN/ANN.h: No such file or directory
In file included from 
build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.cpp:2564:
scikits/ann/kdtree.h:22:21: error: ANN/ANN.h: No such file 
or directory
scikits/ann/kdtree.h:101:3: warning: no newline at end of 
file
scikits/ann/kdtree.h:27: error: ?ANNpointArray? does not 
name a type
scikits/ann/kdtree.h:28: error: ISO C++ forbids 
declaration of ?ANNkd_tree? with no type
scikits/ann/kdtree.h:28: error: expected ?;? before ?*? 
token
scikits/ann/kdtree.h:43: error: ?ANN_KD_SUGGEST? was not 
declared in this scope
scikits/ann/kdtree.h: In constructor 
?_kdtree::_kdtree(double*, int, int, int, int)?:
scikits/ann/kdtree.h:44: error: ?pts? was not declared in 
this scope
scikits/ann/kdtree.h:44: error: ?annAllocPts? was not 
declared in this scope
scikits/ann/kdtree.h:53: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h:53: error: expected type-specifier 
before ?ANNkd_tree?
scikits/ann/kdtree.h:53: error: expected `;' before 
?ANNkd_tree?
scikits/ann/kdtree.h: In destructor ?_kdtree::~_kdtree()?:
scikits/ann/kdtree.h:57: error: ?pts? was not declared in 
this scope
scikits/ann/kdtree.h:57: error: ?annDeallocPts? was not 
declared in this scope
scikits/ann/kdtree.h:58: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h: In member function ?void 
_kdtree::_knn2(double*, int, int, int, int, int*, int, 
int, double*, double) const?:
scikits/ann/kdtree.h:71: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h: In member function ?const char* 
_kdtree::stringRep(bool) const?:
scikits/ann/kdtree.h:98: error: ?tree? was not declared in 
this scope
scikits/ann/kdtree.h:98: error: ?ANNtrue? was not declared 
in this scope
scikits/ann/kdtree.h:98: error: ?ANNfalse? was not 
declared in this scope
error: Command "g++ -pthread -fno-strict-aliasing -DNDEBUG 
-O2 -fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -g -fPIC 
-Iscikits/ann -I/usr/local/include 
-I/usr/local/lib64/python2.5/site-packages/numpy/core/include 
-I/usr/include/python2.5 -c 
build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.cpp -o 
build/temp.linux-x86_64-2.5/build/src.linux-x86_64-2.5/scikits/ann/ANN_wrap.o" 
failed with exit status 1

Nils


From matthieu.brucher at gmail.com  Fri Sep 19 14:23:11 2008
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 19 Sep 2008 20:23:11 +0200
Subject: [SciPy-user] scikits.ann
In-Reply-To: <web-111572639@uni-stuttgart.de>
References: <web-111572639@uni-stuttgart.de>
Message-ID: <e76aa17f0809191123p3a665b3fq2cf99bd67ba0e459@mail.gmail.com>

2008/9/19 Nils Wagner <nwagner at iam.uni-stuttgart.de>:
> Hi all,
>
> I cannot find a component "ann" in the trac
> http://scipy.org/scipy/scikits/
>
> Where can I report a bug ?

Did you install ANN first ?

Matthieu
-- 
French PhD student
Information System Engineer
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 barrywark at gmail.com  Fri Sep 19 16:25:06 2008
From: barrywark at gmail.com (Barry Wark)
Date: Fri, 19 Sep 2008 13:25:06 -0700
Subject: [SciPy-user] scikits.ann
In-Reply-To: <web-111572639@uni-stuttgart.de>
References: <web-111572639@uni-stuttgart.de>
Message-ID: <cd7634ce0809191325pbe4afe4r2e80b898bedd000b@mail.gmail.com>

On Fri, Sep 19, 2008 at 11:01 AM, Nils Wagner
<nwagner at iam.uni-stuttgart.de> wrote:
> Hi all,
>
> I cannot find a component "ann" in the trac
> http://scipy.org/scipy/scikits/
>
> Where can I report a bug ?

Nils,
Some documentation is available at
http://scipy.org/scipy/scikits/wiki/AnnWrapper. Let me know if that
doesn't do the trick for you.

Barry


From nwagner at iam.uni-stuttgart.de  Sat Sep 20 03:28:38 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sat, 20 Sep 2008 09:28:38 +0200
Subject: [SciPy-user] scikits.ann
In-Reply-To: <e76aa17f0809191123p3a665b3fq2cf99bd67ba0e459@mail.gmail.com>
References: <web-111572639@uni-stuttgart.de>
	<e76aa17f0809191123p3a665b3fq2cf99bd67ba0e459@mail.gmail.com>
Message-ID: <web-111588621@uni-stuttgart.de>

On Fri, 19 Sep 2008 20:23:11 +0200
  "Matthieu Brucher" <matthieu.brucher at gmail.com> wrote:
> 2008/9/19 Nils Wagner <nwagner at iam.uni-stuttgart.de>:
>> Hi all,
>>
>> I cannot find a component "ann" in the trac
>> http://scipy.org/scipy/scikits/
>>
>> Where can I report a bug ?
> 
> Did you install ANN first ?
>

Sorry, I missed that. Now it works for me.
  
Thank you very much !

  Nils



From contact at pythonxy.com  Sat Sep 20 11:31:01 2008
From: contact at pythonxy.com (Pierre Raybaut)
Date: Sat, 20 Sep 2008 17:31:01 +0200
Subject: [SciPy-user] [ Python(x,y) ] New release : 2.1.0
Message-ID: <48D51735.4050102@pythonxy.com>

Hi all,

As you may already know, Python(x,y) is a free scientific-oriented
Python Distribution based on Qt and Eclipse providing a self-consistent
scientific development environment.

Release 2.1.0 is now available on http://www.pythonxy.com.
(Full Edition, Basic Edition, Light Edition, Custom Edition and Update)

*Major update*
o Python(x,y) installation is now fully customizable: install only what 
you need! (dependencies are handled by the installer)
(http://www.pythonxy.com/plugins.php)
o Python(x,y) is now designed for (instead of compatible with) Windows 
Vista as well as for Windows XP
(http://www.pythonxy.com/_tools/img.php?lang=&img=//_images/Interactive%20computing.png)

Version 2.1.0  (09-19-2008)

     * Added:
           o Windows installer: fully customizable installation thanks 
to the new plugin-based installer - now you can install only what you 
need among all available Python(x,y) plugins (Python, Eclipse and Others)
     * Updated:
           o xy 1.0.6 (Installed plugins detection)
           o IPython 0.9.1
           o Parallel Python 1.5.6
     * Corrected:
           o Issue 14: Windows Vista / default Eclipse workspace folder 
was incorrect if "My Documents" folder has been moved from its default 
location
           o Issue 16: PyQt4 / plugins not found in Qt Designer

Regards,
Pierre Raybaut


From massimo.sandal at unibo.it  Sun Sep 21 14:46:43 2008
From: massimo.sandal at unibo.it (massimo sandal)
Date: Sun, 21 Sep 2008 20:46:43 +0200
Subject: [SciPy-user] scipy.genetic
In-Reply-To: <1221849174.7186.3.camel@mik>
References: <1221849174.7186.3.camel@mik>
Message-ID: <48D69693.2060809@unibo.it>

Hi Michael,

Michael wrote:
> Hi Massimo,
> 
> thanks for raising this on the list: i have been hunting around for ga
> examples for a while.
> 
> to get hold of the genetic examples:
> 
> 0)	install subversion aka svn
> 1)	download the learn repository 
> 	svn co http://svn.scipy.org/svn/scikits/trunk/learn learn
> the examples are in ~/learn/scikits/learn/machine/ga
> 
> each directory can be individually installed (i think) 
> 
> e.g. maybe by 'python setup.py build' if you want to build it in place
> 
> hope that helps 

Thanks a lot -the localization of the example files was the one I was 
missing.

m.

-- 
Massimo Sandal , Ph.D.
University of Bologna
Department of Biochemistry "G.Moruzzi"

snail mail:
Via Irnerio 48, 40126 Bologna, Italy

email:
massimo.sandal at unibo.it

web:
http://www.biocfarm.unibo.it/samori/people/sandal.html

tel: +39-051-2094388
fax: +39-051-2094387
-------------- next part --------------
A non-text attachment was scrubbed...
Name: massimo_sandal.vcf
Type: text/x-vcard
Size: 274 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080921/fb94a644/attachment.vcf>

From gael.varoquaux at normalesup.org  Mon Sep 22 10:08:34 2008
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 22 Sep 2008 16:08:34 +0200
Subject: [SciPy-user] Proceedings of the SciPy conference.
Message-ID: <20080922140834.GA5840@phare.normalesup.org>

The SciPy conference proceedings are finally available online:
http://conference.scipy.org/proceedings/SciPy2008 .

I hope you enjoy them. I find it great to have this set of excellent
articles talking about works done with, or for, Python in science. For
me, it is a reference to remember what was said at the conference. I hope
it can also be interesting for people who were not present at the
conference.

I apologize for being so slow at publishing them. In addition to the
round trip between authors and editors taking a while, I have been
travelling back home and spent way too much time last week finishing off
administrative duties in the US.

Ga?l


From stefan at sun.ac.za  Mon Sep 22 10:21:43 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Mon, 22 Sep 2008 16:21:43 +0200
Subject: [SciPy-user] Proceedings of the SciPy conference.
In-Reply-To: <20080922140834.GA5840@phare.normalesup.org>
References: <20080922140834.GA5840@phare.normalesup.org>
Message-ID: <9457e7c80809220721k1068f1c4rf0c4a3fde02733ff@mail.gmail.com>

2008/9/22 Gael Varoquaux <gael.varoquaux at normalesup.org>:
> The SciPy conference proceedings are finally available online:
> http://conference.scipy.org/proceedings/SciPy2008 .

Gael, thank you very much for all the time you invested in this
publication.  I know you spent many nights hacking TeX, beating the
processing pipeline into shape, reading, coordinating and encouraging.
 Your efforts did not go by unseen!

Regards
St?fan


From gael.varoquaux at normalesup.org  Mon Sep 22 10:26:40 2008
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Mon, 22 Sep 2008 16:26:40 +0200
Subject: [SciPy-user] Proceedings of the SciPy conference.
In-Reply-To: <9457e7c80809220721k1068f1c4rf0c4a3fde02733ff@mail.gmail.com>
References: <20080922140834.GA5840@phare.normalesup.org>
	<9457e7c80809220721k1068f1c4rf0c4a3fde02733ff@mail.gmail.com>
Message-ID: <20080922142640.GB18576@phare.normalesup.org>

On Mon, Sep 22, 2008 at 04:21:43PM +0200, St?fan van der Walt wrote:
> Gael, thank you very much for all the time you invested in this
> publication.  I know you spent many nights hacking TeX, beating the
> processing pipeline into shape, reading, coordinating and encouraging.
>  Your efforts did not go by unseen!


Thank you for your kind words St?fan. And also thanks for your help with
the webapp. It was a pleasure making the ugliest and most effective hack
of my life with you (sorry for all the gunk I left in there).

Ga?l


From prabhu at aero.iitb.ac.in  Mon Sep 22 10:21:54 2008
From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran)
Date: Mon, 22 Sep 2008 19:51:54 +0530
Subject: [SciPy-user] Proceedings of the SciPy conference.
In-Reply-To: <9457e7c80809220721k1068f1c4rf0c4a3fde02733ff@mail.gmail.com>
References: <20080922140834.GA5840@phare.normalesup.org>
	<9457e7c80809220721k1068f1c4rf0c4a3fde02733ff@mail.gmail.com>
Message-ID: <48D7AA02.9090709@aero.iitb.ac.in>

St?fan van der Walt wrote:
> 2008/9/22 Gael Varoquaux <gael.varoquaux at normalesup.org>:
>> The SciPy conference proceedings are finally available online:
>> http://conference.scipy.org/proceedings/SciPy2008 .
> 
> Gael, thank you very much for all the time you invested in this
> publication.  I know you spent many nights hacking TeX, beating the
> processing pipeline into shape, reading, coordinating and encouraging.
>  Your efforts did not go by unseen!

I second that!  Thanks a ton Ga?l for doing such a great job!

cheers,
prabhu


From Karl.Young at ucsf.edu  Mon Sep 22 12:27:49 2008
From: Karl.Young at ucsf.edu (Karl Young)
Date: Mon, 22 Sep 2008 09:27:49 -0700
Subject: [SciPy-user] Proceedings of the SciPy conference.
In-Reply-To: <48D7AA02.9090709@aero.iitb.ac.in>
References: <20080922140834.GA5840@phare.normalesup.org>	<9457e7c80809220721k1068f1c4rf0c4a3fde02733ff@mail.gmail.com>
	<48D7AA02.9090709@aero.iitb.ac.in>
Message-ID: <48D7C785.5000707@ucsf.edu>

Prabhu Ramachandran wrote:

>St?fan van der Walt wrote:
>  
>
>>2008/9/22 Gael Varoquaux <gael.varoquaux at normalesup.org>:
>>    
>>
>>>The SciPy conference proceedings are finally available online:
>>>http://conference.scipy.org/proceedings/SciPy2008 .
>>>      
>>>
>>Gael, thank you very much for all the time you invested in this
>>publication.  I know you spent many nights hacking TeX, beating the
>>processing pipeline into shape, reading, coordinating and encouraging.
>> Your efforts did not go by unseen!
>>    
>>
>
>I second that!  Thanks a ton Ga?l for doing such a great job!
>
>  
>
I have to chime in too; thanks for the all the hours and all the great 
work Gael; much appreciated !

-- 

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 david at ar.media.kyoto-u.ac.jp  Mon Sep 22 13:37:19 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 23 Sep 2008 02:37:19 +0900
Subject: [SciPy-user] Talkbox 0.1: a scikits for signal processing
 processing with a focus on audio
Message-ID: <48D7D7CF.3030704@ar.media.kyoto-u.ac.jp>

Hi there,

    A quick email to announce the first release of talkbox, a new scikit
of mine. It aims at providing missing features in scipy for signal
processing, with a focus on audio (speech and music mostly). The 0.1 is
mainly a refactoring of some random code I got lying around on my machine:
   
http://www.ar.media.kyoto-u.ac.jp/members/david/softwares/talkbox/index.html

Code available on scikits svn. For this 0.1 release:

    - code for linear prediction coding (LPC) computation: it solves the
Yule Walker equations, both with the straightforward system inversion
and Levinson Durbin.  A C implementation of Levinson Durbin for real,
double autocorrelation is available: it should be reasonably fast
(should be at least comparable to matlab lpc implementation I think).
    - methods for spectral estimation: basic periodogram is implemented.

cheers,

David


From rmay31 at gmail.com  Mon Sep 22 14:36:55 2008
From: rmay31 at gmail.com (Ryan May)
Date: Mon, 22 Sep 2008 13:36:55 -0500
Subject: [SciPy-user] Talkbox 0.1: a scikits for signal processing
 processing with a focus on audio
In-Reply-To: <48D7D7CF.3030704@ar.media.kyoto-u.ac.jp>
References: <48D7D7CF.3030704@ar.media.kyoto-u.ac.jp>
Message-ID: <48D7E5C7.7010006@gmail.com>

David Cournapeau wrote:
> Hi there,
> 
>     A quick email to announce the first release of talkbox, a new scikit
> of mine. It aims at providing missing features in scipy for signal
> processing, with a focus on audio (speech and music mostly). The 0.1 is
> mainly a refactoring of some random code I got lying around on my machine:
>    
> http://www.ar.media.kyoto-u.ac.jp/members/david/softwares/talkbox/index.html
> 
> Code available on scikits svn. For this 0.1 release:
> 
>     - code for linear prediction coding (LPC) computation: it solves the
> Yule Walker equations, both with the straightforward system inversion
> and Levinson Durbin.  A C implementation of Levinson Durbin for real,
> double autocorrelation is available: it should be reasonably fast
> (should be at least comparable to matlab lpc implementation I think).
>     - methods for spectral estimation: basic periodogram is implemented.

Interesting stuff.  Given that some of this is generally quite useful 
across the board (Levinson Durbin, periodogram), I'm hoping that at some 
point it would be possible to push this into mainline SciPy (after a 
period of showing stability and utility)?

Ryan

-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma


From david at ar.media.kyoto-u.ac.jp  Mon Sep 22 14:57:22 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 23 Sep 2008 03:57:22 +0900
Subject: [SciPy-user] Talkbox 0.1: a scikits for signal processing
 processing with a focus on audio
In-Reply-To: <48D7E5C7.7010006@gmail.com>
References: <48D7D7CF.3030704@ar.media.kyoto-u.ac.jp>
	<48D7E5C7.7010006@gmail.com>
Message-ID: <48D7EA92.70708@ar.media.kyoto-u.ac.jp>

Ryan May wrote:
>
> Interesting stuff.  Given that some of this is generally quite useful 
> across the board (Levinson Durbin, periodogram), I'm hoping that at some 
> point it would be possible to push this into mainline SciPy (after a 
> period of showing stability and utility)?
>   

yes, that's one stated goal, and the reason why I use BSD code only (I
could go faster using for example fftw for a lot of dct and co). The
plan was to get 1.0 to implement most of the things I need for signal
processing available in matlab (general), and then for 2.0, more
specialized stuff (MDCT, mfcc, etc...).

cheers,

David


From travis at enthought.com  Mon Sep 22 16:18:17 2008
From: travis at enthought.com (Travis Vaught)
Date: Mon, 22 Sep 2008 15:18:17 -0500
Subject: [SciPy-user] ANN: Enthought Python Distribution 4.0.300 Beta 2
	available
Message-ID: <75D8561B-D264-4F32-AE0B-75CF0C074600@enthought.com>

Greetings,

We've recently posted the second beta release of the Enthought Python  
Distribution (EPD) for our upcoming general release of version 4.0.300  
with Python 2.5.  You may download the beta from here:

http://www.enthought.com/products/epdbeta.php

Please feel free to test it out and provide feedback on the EPD Trac  
instance: https://svn.enthought.com/epd

You can check out the release notes here: http://www.enthought.com/products/epdbetareleasenotes.php


About EPD
---------
The Enthought Python Distribution (EPD) is a "kitchen-sink-included"  
distribution of the Python? Programming Language, including over 60  
additional tools and libraries. The EPD bundle includes NumPy, SciPy,  
IPython, 2D and 3D visualization, database adapters, and a lot of  
other tools right out of the box.

http://www.enthought.com/products/epd.php

It is currently available as a single-click installer for Windows XP  
(x86), Mac OS X (a universal binary for Intel 10.4 and above) and  
RedHat EL3 (x86 and amd64).

EPD is free for academic use.  An annual Subscription and installation  
support are available for individual commercial use (http://www.enthought.com/products/epddownload.php 
).  An Enterprise Subscription with support for particular deployment  
environments is also available for commercial purchase (http://www.enthought.com/products/enterprise.php 
). The Beta versions of EPD are available for indefinite free trial.

Thanks,

Travis

From pitfall66 at freenet.de  Tue Sep 23 08:22:48 2008
From: pitfall66 at freenet.de (Unknown)
Date: Tue, 23 Sep 2008 14:22:48 +0200
Subject: [SciPy-user] vstack / hstack question n dim array without value
Message-ID: <1222172568.13358.11.camel@localhost.whnet>

Hello I am programming this for the Machine Learning Course. I am
studying  Computer Science. 

My Question is can I somehow get rid of the "indices_buffer". Means is
there a easy way to initialize n dim arrays without values ?
I also came up with filling the array with a colum of zeros and deleting
this colum later, but not so elegant in my opinion either.


specifically this code:

indices_buffer=[]

if (indices_buffer==[]):
	indices_buffer=tmp
else:
	indices_buffer=hstack((indices_buffer,tmp))



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

data 
0.3 0.4 0.4
0.2 0.1 0.3
0.3 0.2 0.4
...

cls 
1
1
2
...


def splitcrossval(data,nFold,cls):
	
	indices_buffer=[]
	(rows,columns)=data.shape
	

	classes=unique(cls)
	
	
	for i in classes:

		indices=where(cls==i)[0] 
		
		tmp=hsplit((permutation(indices)),nFold)

		if (indices_buffer==[]):
			indices_buffer=tmp
		else:
			indices_buffer=hstack((indices_buffer,tmp))
		
		
	return 	indices_buffer
	





From sebastian at sipsolutions.net  Tue Sep 23 08:47:15 2008
From: sebastian at sipsolutions.net (Sebastian Stephan Berg)
Date: Tue, 23 Sep 2008 14:47:15 +0200
Subject: [SciPy-user] vstack / hstack question n dim array without value
In-Reply-To: <1222172568.13358.11.camel@localhost.whnet>
References: <1222172568.13358.11.camel@localhost.whnet>
Message-ID: <1222174035.10406.2.camel@sebook>

I think you looking for this?:

In [6]: a = empty((3,0)) # Zero size array ...

In [7]: b = ones((3,27))

In [8]: hstack((a, b)) # Works

Sebastian



From cohen at slac.stanford.edu  Tue Sep 23 10:22:05 2008
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Tue, 23 Sep 2008 16:22:05 +0200
Subject: [SciPy-user] Regarding Genetic algorithm
In-Reply-To: <48D39442.4080905@unibo.it>
References: <7869a7600809180031j282e0221xc21a67b85d540cd5@mail.gmail.com>	<web-111511312@uni-stuttgart.de>
	<48D39442.4080905@unibo.it>
Message-ID: <48D8FB8D.4010408@slac.stanford.edu>

hi, I think that Nils's point is that ga is not in scipy anymore, but in 
scikits, and the track url tells you how to get it : it is in the learn 
package of scikits.
hth,
Johann

massimo sandal wrote:
> Nils Wagner wrote:
>> On Thu, 18 Sep 2008 13:01:44 +0530
>>   "utkarsh wagh" <wagh.utkarsh at gmail.com> wrote:
>>> Hi,
>>>
>>> Can anyone help me out in using the Genetic algorithm (ga) 
>>> subpackage in
>>> Scipy. If possible can anyone send me the sample codes
>>>
>>> thank you,
>>>
>>> -- 
>>> Utkarsh Wagh
>>> IIT Delhi
>>> Contact no: +91 9990646707
>>
>> See
>>
>> http://projects.scipy.org/scipy/scipy/ticket/484
>
> Sorry but I fail to understand how this page can help.
>
> m.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   


From cohen at slac.stanford.edu  Tue Sep 23 10:22:38 2008
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Tue, 23 Sep 2008 16:22:38 +0200
Subject: [SciPy-user] scipy.genetic
In-Reply-To: <48D69693.2060809@unibo.it>
References: <1221849174.7186.3.camel@mik> <48D69693.2060809@unibo.it>
Message-ID: <48D8FBAE.4070607@slac.stanford.edu>

note that the examples.py start with a 'from scipy import ga', which I 
think is obsolete.
Johann

massimo sandal wrote:
> Hi Michael,
>
> Michael wrote:
>> Hi Massimo,
>>
>> thanks for raising this on the list: i have been hunting around for ga
>> examples for a while.
>>
>> to get hold of the genetic examples:
>>
>> 0)    install subversion aka svn
>> 1)    download the learn repository     svn co 
>> http://svn.scipy.org/svn/scikits/trunk/learn learn
>> the examples are in ~/learn/scikits/learn/machine/ga
>>
>> each directory can be individually installed (i think)
>> e.g. maybe by 'python setup.py build' if you want to build it in place
>>
>> hope that helps 
>
> Thanks a lot -the localization of the example files was the one I was 
> missing.
>
> m.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   


From lee.will at gmail.com  Tue Sep 23 10:37:23 2008
From: lee.will at gmail.com (Will Lee)
Date: Tue, 23 Sep 2008 09:37:23 -0500
Subject: [SciPy-user] pyloess in scipy?
Message-ID: <7f03db650809230737s3217e5bfnac836e154a9986c8@mail.gmail.com>

I'm looking to update scipy to something closer to the trunk and I've
discovered that the pyloess.py package under sandbox is missing.  Searching
the mailing list gives me an impression that many modules are now moved to
scikits.  Can somebody point me to where this package is after the
refactoring?

Thanks,

Will
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080923/87afa606/attachment.html>

From williams at astro.ox.ac.uk  Tue Sep 23 10:23:28 2008
From: williams at astro.ox.ac.uk (Michael Williams)
Date: Tue, 23 Sep 2008 15:23:28 +0100
Subject: [SciPy-user] Installation from source on OS X: 'NoneType' object
	has no attribute 'link_shared_object'
Message-ID: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>

Hi,

I'm having trouble installing scipy-0.6.0.tar.gz according to the  
instructions on http://www.scipy.org/Installing_SciPy/Mac_OS_X.

When I run "python setup.py build_src build_clib --fcompiler=gnu95  
build_ext --fcompiler=gnu95 build" I get the following error (email  
continues below):

Traceback (most recent call last):
   File "setup.py", line 53, in <module>
     setup_package()
   File "setup.py", line 45, in setup_package
     configuration=configuration )
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/ 
Extras/lib/python/numpy/distutils/core.py", line 174, in setup
     return old_setup(**new_attr)
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/distutils/core.py", line 151, in setup
     dist.run_commands()
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/distutils/dist.py", line 974, in run_commands
     self.run_command(cmd)
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/distutils/dist.py", line 994, in run_command
     cmd_obj.run()
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/ 
Extras/lib/python/numpy/distutils/command/build_ext.py", line 121, in  
run
     self.build_extensions()
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/distutils/command/build_ext.py", line 416, in build_extensions
     self.build_extension(ext)
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/ 
Extras/lib/python/numpy/distutils/command/build_ext.py", line 312, in  
build_extension
     link = self.fcompiler.link_shared_object
AttributeError: 'NoneType' object has no attribute 'link_shared_object'

Relevant version information follows:

asosx40:~$ gfortran --version
GNU Fortran (GCC) 4.3.0 20070810 (experimental)
asosx40:~$ gcc --version
i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5484)
asosx40:~$ python --version
Python 2.5.1

(This is the stock OS X 10.5 Python installation.)

I get the same error when I try this with the current scipy from svn.

I get a different error when I try to build using g77 by doing "python  
setup.py build_src build_clib --fcompiler=gnu build_ext -- 
fcompiler=gnu build":

creating build/lib.macosx-10.5-i386-2.5/scipy/fftpack
/usr/local/bin/g77 -g -Wall -undefined dynamic_lookup -bundle build/ 
temp.macosx-10.5-i386-2.5/build/src.macosx-10.5-i386-2.5/scipy/fftpack/ 
_fftpackmodule.o build/temp.macosx-10.5-i386-2.5/scipy/fftpack/src/ 
zfft.o build/temp.macosx-10.5-i386-2.5/scipy/fftpack/src/drfft.o build/ 
temp.macosx-10.5-i386-2.5/scipy/fftpack/src/zrfft.o build/ 
temp.macosx-10.5-i386-2.5/scipy/fftpack/src/zfftnd.o build/ 
temp.macosx-10.5-i386-2.5/build/src.macosx-10.5-i386-2.5/ 
fortranobject.o -L/usr/local/lib/gcc/i686-apple-darwin8.8.1/3.4.0 - 
Lbuild/temp.macosx-10.5-i386-2.5 -ldfftpack -lg2c -lcc_dynamic -o  
build/lib.macosx-10.5-i386-2.5/scipy/fftpack/_fftpack.so
ld: library not found for -lcc_dynamic
collect2: ld returned 1 exit status
ld: library not found for -lcc_dynamic
collect2: ld returned 1 exit status
error: Command "/usr/local/bin/g77 -g -Wall -undefined dynamic_lookup - 
bundle build/temp.macosx-10.5-i386-2.5/build/src.macosx-10.5-i386-2.5/ 
scipy/fftpack/_fftpackmodule.o build/temp.macosx-10.5-i386-2.5/scipy/ 
fftpack/src/zfft.o build/temp.macosx-10.5-i386-2.5/scipy/fftpack/src/ 
drfft.o build/temp.macosx-10.5-i386-2.5/scipy/fftpack/src/zrfft.o  
build/temp.macosx-10.5-i386-2.5/scipy/fftpack/src/zfftnd.o build/ 
temp.macosx-10.5-i386-2.5/build/src.macosx-10.5-i386-2.5/ 
fortranobject.o -L/usr/local/lib/gcc/i686-apple-darwin8.8.1/3.4.0 - 
Lbuild/temp.macosx-10.5-i386-2.5 -ldfftpack -lg2c -lcc_dynamic -o  
build/lib.macosx-10.5-i386-2.5/scipy/fftpack/_fftpack.so" failed with  
exit status 1

I would be very grateful for any suggestions!

Thanks,

-- 
Michael Williams
http://www-astro.physics.ox.ac.uk/~williams/


From pgmdevlist at gmail.com  Tue Sep 23 11:09:35 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 23 Sep 2008 11:09:35 -0400
Subject: [SciPy-user] pyloess in scipy?
In-Reply-To: <7f03db650809230737s3217e5bfnac836e154a9986c8@mail.gmail.com>
References: <7f03db650809230737s3217e5bfnac836e154a9986c8@mail.gmail.com>
Message-ID: <200809231109.35988.pgmdevlist@gmail.com>

On Tuesday 23 September 2008 10:37:23 Will Lee wrote:
> I'm looking to update scipy to something closer to the trunk and I've
> discovered that the pyloess.py package under sandbox is missing.  Searching
> the mailing list gives me an impression that many modules are now moved to
> scikits.  Can somebody point me to where this package is after the
> refactoring?

AFAIK, nobody ported pyloess to scikits. I still have the sources of the 
sandbox version lurking on my hard drive. I could send them to you if you're 
interested. I could also try to create a specific scikits, depending on 
whether other people request it (and if it's not a problem for anybody). Let 
me know.




From millman at berkeley.edu  Tue Sep 23 12:45:45 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Tue, 23 Sep 2008 09:45:45 -0700
Subject: [SciPy-user] pyloess in scipy?
In-Reply-To: <200809231109.35988.pgmdevlist@gmail.com>
References: <7f03db650809230737s3217e5bfnac836e154a9986c8@mail.gmail.com>
	<200809231109.35988.pgmdevlist@gmail.com>
Message-ID: <c7009a550809230945r5a53035cha7c68f2cc11d48c1@mail.gmail.com>

On Tue, Sep 23, 2008 at 8:09 AM, Pierre GM <pgmdevlist at gmail.com> wrote:
> On Tuesday 23 September 2008 10:37:23 Will Lee wrote:
>> I'm looking to update scipy to something closer to the trunk and I've
>> discovered that the pyloess.py package under sandbox is missing.  Searching
>> the mailing list gives me an impression that many modules are now moved to
>> scikits.  Can somebody point me to where this package is after the
>> refactoring?
>
> AFAIK, nobody ported pyloess to scikits. I still have the sources of the
> sandbox version lurking on my hard drive. I could send them to you if you're
> interested. I could also try to create a specific scikits, depending on
> whether other people request it (and if it's not a problem for anybody). Let
> me know.

I branched scipy before removing the sandbox:
http://projects.scipy.org/scipy/scipy/browser/branches/sandbox

Here is pyloess:
http://projects.scipy.org/scipy/scipy/browser/branches/sandbox/scipy/sandbox/pyloess

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


From pgmdevlist at gmail.com  Tue Sep 23 13:03:05 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 23 Sep 2008 13:03:05 -0400
Subject: [SciPy-user] pyloess in scipy?
In-Reply-To: <c7009a550809230945r5a53035cha7c68f2cc11d48c1@mail.gmail.com>
References: <7f03db650809230737s3217e5bfnac836e154a9986c8@mail.gmail.com>
	<200809231109.35988.pgmdevlist@gmail.com>
	<c7009a550809230945r5a53035cha7c68f2cc11d48c1@mail.gmail.com>
Message-ID: <200809231303.06494.pgmdevlist@gmail.com>

On Tuesday 23 September 2008 12:45:45 Jarrod Millman wrote:
> Here is pyloess:
> http://projects.scipy.org/scipy/scipy/browser/branches/sandbox/scipy/sandbo
>x/pyloess

Jarrod,
Thanks a lot. Would it be useful to clean up the package and make it a proper 
scikits ?


From robert.kern at gmail.com  Tue Sep 23 13:39:48 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 23 Sep 2008 12:39:48 -0500
Subject: [SciPy-user] Installation from source on OS X: 'NoneType'
	object has no attribute 'link_shared_object'
In-Reply-To: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
Message-ID: <3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>

On Tue, Sep 23, 2008 at 09:23, Michael Williams <williams at astro.ox.ac.uk> wrote:
> Hi,
>
> I'm having trouble installing scipy-0.6.0.tar.gz according to the
> instructions on http://www.scipy.org/Installing_SciPy/Mac_OS_X.
>
> When I run "python setup.py build_src build_clib --fcompiler=gnu95
> build_ext --fcompiler=gnu95 build" I get the following error (email
> continues below):

Can you show us more of the build log? Particularly the parts during
the Fortran compiler detection. What version of numpy are you using?

> I get a different error when I try to build using g77 by doing "python
> setup.py build_src build_clib --fcompiler=gnu build_ext --
> fcompiler=gnu build":

Yeah, you just can't use g77 with a Universal Python.

-- 
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 williams at astro.ox.ac.uk  Tue Sep 23 14:22:37 2008
From: williams at astro.ox.ac.uk (Michael Williams)
Date: Tue, 23 Sep 2008 19:22:37 +0100
Subject: [SciPy-user] Installation from source on OS X: 'NoneType'
	object has no attribute 'link_shared_object'
In-Reply-To: <3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
	<3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
Message-ID: <BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>

Hi Robert,

thanks very much.

On 23 Sep 2008, at 18:39, Robert Kern wrote:
> Can you show us more of the build log? Particularly the parts during
> the Fortran compiler detection. What version of numpy are you using?

numpy is version 1.1.1. I'm very confused because I'm getting a  
different error message when I install scipy. I'm pretty sure I  
haven't done anything that has changed my setup. To confirm that this  
new error message is real I removed my site-packages directory and ran  
"easy_install numpy".

The last 100 lines of the build log, which seem to include the  
detection of the Fortran compiler, are below. The full log is here: http://rafb.net/p/Sue0Wv11.html 
.

Hope this helps. Thanks,
-- Mike

running build_clib
customize UnixCCompiler
customize UnixCCompiler using build_clib
customize Gnu95FCompiler
Found executable /usr/local/bin/gfortran
customize Gnu95FCompiler using build_clib
building 'dfftpack' library
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -fPIC -O3 -funroll-loops
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second- 
underscore -fPIC -O3 -funroll-loops
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -Wall -fno-second-underscore -fPIC -O3 -funroll-loops
creating build/temp.macosx-10.5-i386-2.5
creating build/temp.macosx-10.5-i386-2.5/scipy
creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack
creating build/temp.macosx-10.5-i386-2.5/scipy/fftpack/dfftpack
compile options: '-c'
gfortran:f77: scipy/fftpack/dfftpack/dcosqb.f
gfortran:f77: scipy/fftpack/dfftpack/dcosqf.f
gfortran:f77: scipy/fftpack/dfftpack/dcosqi.f
gfortran:f77: scipy/fftpack/dfftpack/dcost.f
gfortran:f77: scipy/fftpack/dfftpack/dcosti.f
gfortran:f77: scipy/fftpack/dfftpack/dfftb.f
gfortran:f77: scipy/fftpack/dfftpack/dfftb1.f
gfortran:f77: scipy/fftpack/dfftpack/dfftf.f
gfortran:f77: scipy/fftpack/dfftpack/dfftf1.f
gfortran:f77: scipy/fftpack/dfftpack/dffti.f
gfortran:f77: scipy/fftpack/dfftpack/dffti1.f
scipy/fftpack/dfftpack/dffti1.f: In function ?dffti1?:
scipy/fftpack/dfftpack/dffti1.f:11: warning: ?ntry? may be used  
uninitialized in this function
gfortran:f77: scipy/fftpack/dfftpack/dsinqb.f
gfortran:f77: scipy/fftpack/dfftpack/dsinqf.f
gfortran:f77: scipy/fftpack/dfftpack/dsinqi.f
gfortran:f77: scipy/fftpack/dfftpack/dsint.f
gfortran:f77: scipy/fftpack/dfftpack/dsint1.f
gfortran:f77: scipy/fftpack/dfftpack/dsinti.f
gfortran:f77: scipy/fftpack/dfftpack/zfftb.f
gfortran:f77: scipy/fftpack/dfftpack/zfftb1.f
gfortran:f77: scipy/fftpack/dfftpack/zfftf.f
gfortran:f77: scipy/fftpack/dfftpack/zfftf1.f
gfortran:f77: scipy/fftpack/dfftpack/zffti.f
gfortran:f77: scipy/fftpack/dfftpack/zffti1.f
scipy/fftpack/dfftpack/zffti1.f: In function ?zffti1?:
scipy/fftpack/dfftpack/zffti1.f:11: warning: ?ntry? may be used  
uninitialized in this function
ar: adding 23 object files to build/temp.macosx-10.5-i386-2.5/ 
libdfftpack.a
ranlib:@ build/temp.macosx-10.5-i386-2.5/libdfftpack.a
building 'linpack_lite' library
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -fPIC -O3 -funroll-loops
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second- 
underscore -fPIC -O3 -funroll-loops
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -Wall -fno-second-underscore -fPIC -O3 -funroll-loops
creating build/temp.macosx-10.5-i386-2.5/scipy/integrate
creating build/temp.macosx-10.5-i386-2.5/scipy/integrate/linpack_lite
compile options: '-c'
gfortran:f77: scipy/integrate/linpack_lite/dgbfa.f
gfortran:f77: scipy/integrate/linpack_lite/dgbsl.f
gfortran:f77: scipy/integrate/linpack_lite/dgefa.f
gfortran:f77: scipy/integrate/linpack_lite/dgesl.f
gfortran:f77: scipy/integrate/linpack_lite/dgtsl.f
ar: adding 5 object files to build/temp.macosx-10.5-i386-2.5/ 
liblinpack_lite.a
ranlib:@ build/temp.macosx-10.5-i386-2.5/liblinpack_lite.a
building 'mach' library
using additional config_fc from setup script for fortran compiler:  
{'noopt': ('scipy/integrate/setup.py', 1)}
customize Gnu95FCompiler
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -fPIC
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second- 
underscore -fPIC
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -Wall -fno-second-underscore -fPIC
creating build/temp.macosx-10.5-i386-2.5/scipy/integrate/mach
compile options: '-c'
gfortran:f77: scipy/integrate/mach/d1mach.f
gfortran:f77: scipy/integrate/mach/i1mach.f
gfortran:f77: scipy/integrate/mach/r1mach.f
gfortran:f77: scipy/integrate/mach/xerror.f
scipy/integrate/mach/xerror.f:1.40:

       SUBROUTINE XERROR(MESS,NMESS,L1,L2)
                                        1
Warning: Unused dummy argument 'l2' at (1)
scipy/integrate/mach/xerror.f:1.37:

       SUBROUTINE XERROR(MESS,NMESS,L1,L2)
                                     1
Warning: Unused dummy argument 'l1' at (1)
ar: adding 4 object files to build/temp.macosx-10.5-i386-2.5/libmach.a
ranlib:@ build/temp.macosx-10.5-i386-2.5/libmach.a
building 'quadpack' library
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -fPIC -O3 -funroll-loops
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second- 
underscore -fPIC -O3 -funroll-loops
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form -fno- 
second-underscore -Wall -fno-second-underscore -fPIC -O3 -funroll-loops
creating build/temp.macosx-10.5-i386-2.5/scipy/integrate/quadpack
compile options: '-c'
gfortran:f77: scipy/integrate/quadpack/dqag.f
gfortran:f77: scipy/integrate/quadpack/dqage.f
gfortran:f77: scipy/integrate/quadpack/dqagi.f
gfortran:f77: scipy/integrate/quadpack/dqagie.f
scipy/integrate/quadpack/dqagie.f: In function ?dqagie?:
scipy/integrate/quadpack/dqagie.f:154: warning: ?small? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqagie.f:153: warning: ?ertest? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqagie.f:152: warning: ?erlarg? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqagie.f:151: warning: ?correc? may be used  
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqagp.f
gfortran:f77: scipy/integrate/quadpack/dqagpe.f
scipy/integrate/quadpack/dqagpe.f: In function ?dqagpe?:
scipy/integrate/quadpack/dqagpe.f:196: warning: ?k? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqagpe.f:191: warning: ?correc? may be used  
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqags.f
gfortran:f77: scipy/integrate/quadpack/dqagse.f
scipy/integrate/quadpack/dqagse.f: In function ?dqagse?:
scipy/integrate/quadpack/dqagse.f:153: warning: ?small? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqagse.f:152: warning: ?ertest? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqagse.f:151: warning: ?erlarg? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqagse.f:150: warning: ?correc? may be used  
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqawc.f
gfortran:f77: scipy/integrate/quadpack/dqawce.f
gfortran:f77: scipy/integrate/quadpack/dqawf.f
gfortran:f77: scipy/integrate/quadpack/dqawfe.f
scipy/integrate/quadpack/dqawfe.f: In function ?dqawfe?:
scipy/integrate/quadpack/dqawfe.f:203: warning: ?ll? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqawfe.f:200: warning: ?drl? may be used  
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqawo.f
gfortran:f77: scipy/integrate/quadpack/dqawoe.f
scipy/integrate/quadpack/dqawoe.f: In function ?dqawoe?:
scipy/integrate/quadpack/dqawoe.f:208: warning: ?ertest? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqawoe.f:207: warning: ?erlarg? may be used  
uninitialized in this function
scipy/integrate/quadpack/dqawoe.f:206: warning: ?correc? may be used  
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqaws.f
gfortran:f77: scipy/integrate/quadpack/dqawse.f
gfortran:f77: scipy/integrate/quadpack/dqc25c.f
gfortran:f77: scipy/integrate/quadpack/dqc25f.f
scipy/integrate/quadpack/dqc25f.f: In function ?dqc25f?:
scipy/integrate/quadpack/dqc25f.f:103: warning: ?m? may be used  
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqc25s.f
gfortran:f77: scipy/integrate/quadpack/dqcheb.f
gfortran:f77: scipy/integrate/quadpack/dqelg.f
scipy/integrate/quadpack/dqelg.f: In function ?dqelg?:
scipy/integrate/quadpack/dqelg.f:1: internal compiler error: vector  
VEC(tree,base) index domain error, in build_classic_dist_vector_1 at  
tree-data-ref.c:2725
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
scipy/integrate/quadpack/dqelg.f: In function ?dqelg?:
scipy/integrate/quadpack/dqelg.f:1: internal compiler error: vector  
VEC(tree,base) index domain error, in build_classic_dist_vector_1 at  
tree-data-ref.c:2725
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
error: Command "/usr/local/bin/gfortran -Wall -ffixed-form -fno-second- 
underscore -fPIC -O3 -funroll-loops -c -c scipy/integrate/quadpack/ 
dqelg.f -o build/temp.macosx-10.5-i386-2.5/scipy/integrate/quadpack/ 
dqelg.o" failed with exit status 1





From robert.kern at gmail.com  Tue Sep 23 14:53:21 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 23 Sep 2008 13:53:21 -0500
Subject: [SciPy-user] Installation from source on OS X: 'NoneType'
	object has no attribute 'link_shared_object'
In-Reply-To: <BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
	<3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
	<BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>
Message-ID: <3d375d730809231153p43290a4bu10cdf119908c8830@mail.gmail.com>

On Tue, Sep 23, 2008 at 13:22, Michael Williams <williams at astro.ox.ac.uk> wrote:
> Hi Robert,
>
> thanks very much.
>
> On 23 Sep 2008, at 18:39, Robert Kern wrote:
>> Can you show us more of the build log? Particularly the parts during
>> the Fortran compiler detection. What version of numpy are you using?
>
> numpy is version 1.1.1. I'm very confused because I'm getting a
> different error message when I install scipy. I'm pretty sure I
> haven't done anything that has changed my setup. To confirm that this
> new error message is real I removed my site-packages directory and ran
> "easy_install numpy".
>
> The last 100 lines of the build log, which seem to include the
> detection of the Fortran compiler, are below. The full log is here: http://rafb.net/p/Sue0Wv11.html
> .

Unfortunately, now it looks like you are dealing with a bug in
gfortran. Can you downgrade to something earlier than 4.3.0?

-- 
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 andreas.matthias at gmail.com  Tue Sep 23 16:46:16 2008
From: andreas.matthias at gmail.com (Andreas Matthias)
Date: Tue, 23 Sep 2008 22:46:16 +0200
Subject: [SciPy-user] reading character arrays from fortran file
Message-ID: <o51oq5-a2f.ln1@fawkes.hogwarts>

I'm trying to read a character array from a binary fortran file
and the following code seems to work:

from scipy.io import fopen
fd1 = fopen.fopen('fort', permission='rb', format='n')
print fd1.fort_read(0, dtype='str')

But then I get this deprecation warning saying that I should
use npfile instead of fopen. Unfortunately, I don't get npfile
to do the same as the code above. I tried the following which
does not work:

from scipy.io import npfile
fd2 = npfile('fort')
print fd2.read_array(dt='str', shape=-1)

What's the correct way to do it with npfile?


Ciao
Andreas



From kdsudac at yahoo.com  Tue Sep 23 18:28:56 2008
From: kdsudac at yahoo.com (Keith Suda-Cederquist)
Date: Tue, 23 Sep 2008 15:28:56 -0700 (PDT)
Subject: [SciPy-user] Pickling Large (Image) Arrays
Message-ID: <568345.52718.qm@web54305.mail.re2.yahoo.com>

Hi All,

I'm doing some image processing on some rather large images (2000x2000 pixels and each pixel has 16 bits) so the file comes in at 7-8 MB.  During the image processing I convert the image to a 64-bit float numpy array and do a bunch of operations on the image.

In certain cases (where tests fail), I'd like to save all the data to a file to take a look at later and debug.  I need to keep the size of this file as small as possible.

I'm thinking of writing some code that will round pixel values to an 8-bit unsigned integer and then pickle the data to a file.  Is this the a good approach?  Can anyone suggest a better approach?  Will this actually succeed in reducing the file size, or will I just be wasting my time?

Thanks,
Keith



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080923/58ee7833/attachment.html>

From peridot.faceted at gmail.com  Tue Sep 23 19:13:16 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Tue, 23 Sep 2008 19:13:16 -0400
Subject: [SciPy-user] Pickling Large (Image) Arrays
In-Reply-To: <568345.52718.qm@web54305.mail.re2.yahoo.com>
References: <568345.52718.qm@web54305.mail.re2.yahoo.com>
Message-ID: <ce557a360809231613u3fafed33h9f11d357412176ac@mail.gmail.com>

2008/9/23 Keith Suda-Cederquist <kdsudac at yahoo.com>:

> I'm doing some image processing on some rather large images (2000x2000
> pixels and each pixel has 16 bits) so the file comes in at 7-8 MB.  During
> the image processing I convert the image to a 64-bit float numpy array and
> do a bunch of operations on the image.
>
> In certain cases (where tests fail), I'd like to save all the data to a file
> to take a look at later and debug.  I need to keep the size of this file as
> small as possible.
>
> I'm thinking of writing some code that will round pixel values to an 8-bit
> unsigned integer and then pickle the data to a file.  Is this the a good
> approach?  Can anyone suggest a better approach?  Will this actually succeed
> in reducing the file size, or will I just be wasting my time?

First of all, the current release of numpy includes a native file
format that is fairly efficient, fast, and portable. If you can, it's
probably better to use that than pickles. But by itself it won't save
all that much space: almost all the space in either format is taken up
by the pixel array. If you convert the pixel array to one with dtype
uint8 or unit16, you'll use one byte per pixel instead of eight. You
do of course lose information this way, and if this obscures why the
test is failing, it will be quite frustrating. If the data is very
compressible, you could look into using the Python Imaging Library to
save it in some compressed image format, though this will almost
certainly lose even more information.

Anne


From mscipy at googlemail.com  Wed Sep 24 03:57:43 2008
From: mscipy at googlemail.com (Saber Mbarek)
Date: Wed, 24 Sep 2008 07:57:43 +0000
Subject: [SciPy-user] importing pylab??
Message-ID: <90ec20e0809240057h64f91941t6a246678a098be43@mail.gmail.com>

Hi,

I installed scipy a few days ago on my laptop (with os linux debian), but
while importing the pylab module from the (installed) python-matplotlib I
got the following error:

-------------------------------------------------------------------------------------------------------------
Python 2.5.2 (r252:60911, Aug  8 2008, 09:22:44)
[GCC 4.3.1] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import scipy
>>> import numpy
>>> import pylab
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python2.5/site-packages/pylab.py", line 1, in <module>
    from matplotlib.pylab import *
  File "/usr/lib/python2.5/site-packages/matplotlib/__init__.py", line 128,
in <module>
    from rcsetup import defaultParams, validate_backend, validate_toolbar
  File "/usr/lib/python2.5/site-packages/matplotlib/rcsetup.py", line 18, in
<module>
    from matplotlib.colors import is_color_like
  File "/usr/lib/python2.5/site-packages/matplotlib/colors.py", line 39, in
<module>
    import matplotlib.cbook as cbook
  File "/usr/lib/python2.5/site-packages/matplotlib/cbook.py", line 14, in
<module>
    preferredencoding = locale.getpreferredencoding()
  File "/usr/lib/python2.5/locale.py", line 514, in getpreferredencoding
    setlocale(LC_CTYPE, "")
  File "/usr/lib/python2.5/locale.py", line 478, in setlocale
    return _setlocale(category, locale)
locale.Error: unsupported locale setting
------------------------------------------------------------------------------------------------------------

Could you please help me ?

Best regards,
Saber
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080924/dbee4ab6/attachment.html>

From robert.kern at gmail.com  Wed Sep 24 04:14:20 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 24 Sep 2008 03:14:20 -0500
Subject: [SciPy-user] importing pylab??
In-Reply-To: <90ec20e0809240057h64f91941t6a246678a098be43@mail.gmail.com>
References: <90ec20e0809240057h64f91941t6a246678a098be43@mail.gmail.com>
Message-ID: <3d375d730809240114m65fc8195n9857ab4c6d590395@mail.gmail.com>

On Wed, Sep 24, 2008 at 02:57, Saber Mbarek <mscipy at googlemail.com> wrote:
> Hi,
>
> I installed scipy a few days ago on my laptop (with os linux debian), but
> while importing the pylab module from the (installed) python-matplotlib I
> got the following error:

You want the matplotlib list over here:

  https://lists.sourceforge.net/lists/listinfo/matplotlib-users

-- 
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 jr at sun.ac.za  Wed Sep 24 10:05:01 2008
From: jr at sun.ac.za (Johann Rohwer)
Date: Wed, 24 Sep 2008 16:05:01 +0200
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
Message-ID: <200809241605.01575.jr@sun.ac.za>

Hi 

It seems that the default implementation of std and var differs 
between numpy/scipy and scipy.stats, in that numpy/scipy is using 
the "biased" formulation (i.e. dividing by N) whereas scipy.stats is 
using the "unbiased" formulation (dividing by N-1) by default. Is 
this intentional (it could be potentially confusing...)? I realise 
that the "biased" version can be accessed in sp.stats with a kwarg, 
but what is the reason for two different implementations of the 
function(s)?

In [30]: a
Out[30]: array([ 1.,  2.,  3.,  2.,  3.,  1.])

In [31]: np.std(a)
Out[31]: 0.81649658092772603

In [32]: sp.std(a)
Out[32]: 0.81649658092772603

In [33]: sp.stats.std(a)
Out[33]: 0.89442719099991586

In [34]: sp.stats.std(a, bias=True)
Out[34]: 0.81649658092772603

Same for np.var vs scipy.stats.var

Johann


From pgmdevlist at gmail.com  Wed Sep 24 12:05:02 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Wed, 24 Sep 2008 12:05:02 -0400
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <200809241605.01575.jr@sun.ac.za>
References: <200809241605.01575.jr@sun.ac.za>
Message-ID: <200809241205.02637.pgmdevlist@gmail.com>

Johann,
You can also get the unbiased estimates with numpy by setting the optional 
parameter ddof=1.

>>> a=np.array([ 1., ?2., ?3., ?2., ?3., ?1.])
>>> np.std(a)
0.81649658092772603
>>> np.std(a, ddof=1)
0.89442719099991586

I think the default to biased estimates was kept for backward compatibility.



From josef.pktd at gmail.com  Wed Sep 24 12:19:37 2008
From: josef.pktd at gmail.com (joep)
Date: Wed, 24 Sep 2008 09:19:37 -0700 (PDT)
Subject: [SciPy-user] .typecode in scipy.stats
Message-ID: <9e6ca619-3ec2-415e-ac18-5bca546a81f6@c58g2000hsc.googlegroups.com>

``.typecode`` seems to be depreciated for a long time, but there is
still one left in \scipy\stats\_support.py (is still in current trunk)

see traceback

Josef

>>> stats.itemfreq(freq)
Traceback (most recent call last):
  File "<pyshell#12>", line 1, in ?
    stats.itemfreq(freq)
  File "c:\programs\python24\lib\site-packages\scipy\stats\stats.py",
line 954, in itemfreq
    scores = _support.unique(a)
  File "c:\programs\python24\lib\site-packages\scipy\stats
\_support.py", line 51, in unique
    if inarray.typecode() != 'O':  # not an Object array
AttributeError: 'numpy.ndarray' object has no attribute 'typecode'


From oliphant at enthought.com  Wed Sep 24 13:18:17 2008
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Wed, 24 Sep 2008 12:18:17 -0500
Subject: [SciPy-user] .typecode in scipy.stats
In-Reply-To: <9e6ca619-3ec2-415e-ac18-5bca546a81f6@c58g2000hsc.googlegroups.com>
References: <9e6ca619-3ec2-415e-ac18-5bca546a81f6@c58g2000hsc.googlegroups.com>
Message-ID: <48DA7659.4070804@enthought.com>

joep wrote:
> ``.typecode`` seems to be depreciated for a long time, but there is
> still one left in \scipy\stats\_support.py (is still in current trunk)
>
>   
Thanks for the bug report.   Fixed in r4741.

-Travis



From aisaac at american.edu  Wed Sep 24 15:29:29 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Wed, 24 Sep 2008 15:29:29 -0400
Subject: [SciPy-user] Inconsistent standard deviation and
 variance	implementation in scipy vs. scipy.stats
In-Reply-To: <200809241205.02637.pgmdevlist@gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
Message-ID: <48DA9519.1000203@american.edu>

On 9/24/2008 12:05 PM Pierre GM apparently wrote:
> I think the default to biased estimates was kept for backward compatibility.

It is still a problem that scip.var and scipy.stats.var
behave differently (and even have a different signature).
What is the way forward?

An opening suggestion:
unify the signature, let ``bias`` be a deprecated way to
to set ``ddof``, and warn users of scipy.stats.var (or std)
if they do not set ``ddof``.

Alan Isaac



From bjracine at glosten.com  Wed Sep 24 16:00:30 2008
From: bjracine at glosten.com (Benjamin J. Racine)
Date: Wed, 24 Sep 2008 13:00:30 -0700
Subject: [SciPy-user] Quick Reference Card
Message-ID: <8C2B20C4348091499673D86BF10AB6761ACA8097@clipper.glosten.local>

I'd like to make a python/numpy/scipy equivalent of the MATLAB quick reference:
http://www.math.umd.edu/~jeo/matlab_quickref.pdf
I think that this could be an improvement upon:
http://mathesaurus.sourceforge.net/matlab-python-xref.pdf
and go a long ways in putting newcomers on their feet.
Any thoughts or advice?
Ben R.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080924/8a6065c0/attachment.html>

From gael.varoquaux at normalesup.org  Wed Sep 24 16:13:14 2008
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Wed, 24 Sep 2008 22:13:14 +0200
Subject: [SciPy-user] Quick Reference Card
In-Reply-To: <8C2B20C4348091499673D86BF10AB6761ACA8097@clipper.glosten.local>
References: <8C2B20C4348091499673D86BF10AB6761ACA8097@clipper.glosten.local>
Message-ID: <20080924201314.GJ7690@phare.normalesup.org>

On Wed, Sep 24, 2008 at 01:00:30PM -0700, Benjamin J. Racine wrote:
>    I'd like to make a python/numpy/scipy equivalent of the MATLAB quick
>    reference:

>    [1]http://www.math.umd.edu/~jeo/matlab_quickref.pdf

>    I think that this could be an improvement upon:

>    [2]http://mathesaurus.sourceforge.net/matlab-python-xref.pdf

>    and go a long ways in putting newcomers on their feet.

>    Any thoughts or advice?

Go for it! It would be great. And if you can link it from the wiki,
people are likely to find it, and appreciate it.

Ga?l


From peridot.faceted at gmail.com  Wed Sep 24 16:35:45 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Wed, 24 Sep 2008 16:35:45 -0400
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <48DA9519.1000203@american.edu>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
Message-ID: <ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>

2008/9/24 Alan G Isaac <aisaac at american.edu>:
> On 9/24/2008 12:05 PM Pierre GM apparently wrote:
>> I think the default to biased estimates was kept for backward compatibility.
>
> It is still a problem that scip.var and scipy.stats.var
> behave differently (and even have a different signature).
> What is the way forward?
>
> An opening suggestion:
> unify the signature, let ``bias`` be a deprecated way to
> to set ``ddof``, and warn users of scipy.stats.var (or std)
> if they do not set ``ddof``.

How about (possibly in addition to your suggestion) deprecating the
re-exporting of numpy functions inside scipy? People often seem to ask
about whether they should be using the scipy "version" or the numpy
"version" of some function, when in fact it's just a re-exporting of
the name.

This still leaves the question of an inconsistency between scipy and
numpy, for which I think your suggestion is a reasonable solution.

Anne


From stef.mientki at gmail.com  Wed Sep 24 17:11:31 2008
From: stef.mientki at gmail.com (Stef Mientki)
Date: Wed, 24 Sep 2008 23:11:31 +0200
Subject: [SciPy-user] Quick Reference Card
In-Reply-To: <20080924201314.GJ7690@phare.normalesup.org>
References: <8C2B20C4348091499673D86BF10AB6761ACA8097@clipper.glosten.local>
	<20080924201314.GJ7690@phare.normalesup.org>
Message-ID: <48DAAD03.10906@gmail.com>


>>    Any thoughts or advice?
>>     
>
> Go for it! It would be great. And if you can link it from the wiki,
> people are likely to find it, and appreciate it.
>   
+1

(btw I like the Python 2.4 reference card of Laurent Pointal a lot:
http://www.digilife.be/quickreferences/QRC/Python%202.4%20Quick%20Reference%20Card.pdf
on the other hand for a complex package as Scipy++ it would be more 
valuable,
if more than 1 person could contribute.

cheers,
Stef



From matthew.brett at gmail.com  Wed Sep 24 20:20:44 2008
From: matthew.brett at gmail.com (Matthew Brett)
Date: Wed, 24 Sep 2008 17:20:44 -0700
Subject: [SciPy-user] pyloess in scipy?
In-Reply-To: <200809231303.06494.pgmdevlist@gmail.com>
References: <7f03db650809230737s3217e5bfnac836e154a9986c8@mail.gmail.com>
	<200809231109.35988.pgmdevlist@gmail.com>
	<c7009a550809230945r5a53035cha7c68f2cc11d48c1@mail.gmail.com>
	<200809231303.06494.pgmdevlist@gmail.com>
Message-ID: <1e2af89e0809241720j7522933asf7042aa69e0aa944@mail.gmail.com>

Hi,

On Tue, Sep 23, 2008 at 10:03 AM, Pierre GM <pgmdevlist at gmail.com> wrote:
> On Tuesday 23 September 2008 12:45:45 Jarrod Millman wrote:
>> Here is pyloess:
>> http://projects.scipy.org/scipy/scipy/browser/branches/sandbox/scipy/sandbo
>>x/pyloess
>
> Jarrod,
> Thanks a lot. Would it be useful to clean up the package and make it a proper
> scikits ?

I'm attaching a slightly modified version of the pure python biopython
implementation.

However, I don't know whether it (or the original) works correctly.

This in hope that someone who knows this stuff better than me is interested.

Biopython implementation here:

http://cvs.biopython.org/cgi-bin/viewcvs/viewcvs.cgi/biopython/Bio/Statistics/lowess.py?cvsroot=biopython

I was expecting this to work:

n = 100
x = np.arange(n)
y = np.random.normal(size=(n,))
yest = lowess(x, y)

but it returns a LinAlgError (as does the unmodified original).

I see that R has a more complex algorithm, based on the original
fortran.  For reference, I have attached the file (lowess.doc) that
they worked from, that used to be in the R source.

Any thoughts?

Matthew
-------------- next part --------------
A non-text attachment was scrubbed...
Name: lowess.doc
Type: application/msword
Size: 20671 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080924/6f6d1883/attachment.doc>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: lowess.py
Type: text/x-python
Size: 3444 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080924/6f6d1883/attachment.py>

From cournape at gmail.com  Thu Sep 25 00:35:18 2008
From: cournape at gmail.com (David Cournapeau)
Date: Thu, 25 Sep 2008 13:35:18 +0900
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <48DA9519.1000203@american.edu>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
Message-ID: <5b8d13220809242135w7174b0f3l81cf81bddbd3ed89@mail.gmail.com>

On Thu, Sep 25, 2008 at 4:29 AM, Alan G Isaac <aisaac at american.edu> wrote:
> An opening suggestion:
> unify the signature, let ``bias`` be a deprecated way to
> to set ``ddof``, and warn users of scipy.stats.var (or std)
> if they do not set ``ddof``.

The problem is that there is another discrepancy between numpy.var and
scipy.stats.var: the axis argument is 0 in scipy.stats, None in numpy.
So if we want to be really compatible, we can't just add a new
argument and deprecate the old one; we have to deprecate the current
signature, and change it later.

I suggested some time ago to deprecate scipy.stats current signature
for 0.7, and set the new one in 0.8. If that's fine with you, we could
do that. I don't feel confortable changing a function in scipy.stats
(because it is not "my" module), but OTOH, nobody reacted last time we
had this discussion, so maybe we should just do it.

David


From stefan at sun.ac.za  Thu Sep 25 03:48:36 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Thu, 25 Sep 2008 09:48:36 +0200
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
Message-ID: <9457e7c80809250048r5bec3240x35d25c8c791ee728@mail.gmail.com>

2008/9/24 Anne Archibald <peridot.faceted at gmail.com>:
> How about (possibly in addition to your suggestion) deprecating the
> re-exporting of numpy functions inside scipy? People often seem to ask
> about whether they should be using the scipy "version" or the numpy
> "version" of some function, when in fact it's just a re-exporting of
> the name.

I'm all in favour of that suggestion.

Cheers
St?fan


From stefan at sun.ac.za  Thu Sep 25 03:50:28 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Thu, 25 Sep 2008 09:50:28 +0200
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <5b8d13220809242135w7174b0f3l81cf81bddbd3ed89@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<5b8d13220809242135w7174b0f3l81cf81bddbd3ed89@mail.gmail.com>
Message-ID: <9457e7c80809250050u15ca8517o727f87936236da78@mail.gmail.com>

2008/9/25 David Cournapeau <cournape at gmail.com>:
> I suggested some time ago to deprecate scipy.stats current signature
> for 0.7, and set the new one in 0.8. If that's fine with you, we could
> do that. I don't feel confortable changing a function in scipy.stats
> (because it is not "my" module), but OTOH, nobody reacted last time we
> had this discussion, so maybe we should just do it.

Yes, do it.  If someone complains, that's why we have SVN.  The SciPy
API is not (and cannot be) frozen -- it is still too immature, so
let's get it up to scratch ASAP.

Cheers
St?fan


From pav at iki.fi  Thu Sep 25 04:38:19 2008
From: pav at iki.fi (Pauli Virtanen)
Date: Thu, 25 Sep 2008 08:38:19 +0000 (UTC)
Subject: [SciPy-user] Inconsistent standard deviation and
	variance	implementation in scipy vs. scipy.stats
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
Message-ID: <gbfilr$vls$1@ger.gmane.org>

Wed, 24 Sep 2008 16:35:45 -0400, Anne Archibald wrote:
> 2008/9/24 Alan G Isaac <aisaac at american.edu>:
>> On 9/24/2008 12:05 PM Pierre GM apparently wrote:
>>> I think the default to biased estimates was kept for backward
>>> compatibility.
>>
>> It is still a problem that scip.var and scipy.stats.var behave
>> differently (and even have a different signature). What is the way
>> forward?
>>
>> An opening suggestion:
>> unify the signature, let ``bias`` be a deprecated way to to set
>> ``ddof``, and warn users of scipy.stats.var (or std) if they do not set
>> ``ddof``.
> 
> How about (possibly in addition to your suggestion) deprecating the
> re-exporting of numpy functions inside scipy? People often seem to ask
> about whether they should be using the scipy "version" or the numpy
> "version" of some function, when in fact it's just a re-exporting of the
> name.

The opposite direction would be completely removing `var` from 
scipy.stats. Is there a reason why the function is reimplemented in 
scipy? There's probably need eg. for float -> complex casting sqrt(), but 
I don't clearly see why there are two variants of `var`.

Personally, I'd prefer not to have the same function reimplemented in two 
places, unless there is a clear need for it. I think there are more 
examples of duplication / signature mismatches in scipy vs. numpy that 
could be cleaned up a bit, at least in scipy.linalg.

-- 
Pauli Virtanen



From jr at sun.ac.za  Thu Sep 25 07:39:10 2008
From: jr at sun.ac.za (Johann Rohwer)
Date: Thu, 25 Sep 2008 13:39:10 +0200
Subject: [SciPy-user]
	=?iso-8859-1?q?Inconsistent_standard_deviation_and_v?=
	=?iso-8859-1?q?ariance=09implementation_in_scipy_vs=2E_scipy=2Esta?=
	=?iso-8859-1?q?ts?=
In-Reply-To: <gbfilr$vls$1@ger.gmane.org>
References: <200809241605.01575.jr@sun.ac.za>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
	<gbfilr$vls$1@ger.gmane.org>
Message-ID: <200809251339.10790.jr@sun.ac.za>

On Thursday, 25 September 2008, Pauli Virtanen wrote:
> The opposite direction would be completely removing `var` from
> scipy.stats. Is there a reason why the function is reimplemented in
> scipy? There's probably need eg. for float -> complex casting
> sqrt(), but I don't clearly see why there are two variants of
> `var`.
>
> Personally, I'd prefer not to have the same function reimplemented
> in two places, unless there is a clear need for it. I think there
> are more examples of duplication / signature mismatches in scipy
> vs. numpy that could be cleaned up a bit, at least in scipy.linalg.

I agree that duplicate implementations of the same function are 
confusing. 

However, within numpy itself there is further inconsistency, in that 
np.var and np.std use the "ddof" kwarg, whereas np.cov uses 
the "bias" kwarg (as do sp.stats.std and sp.stats.var). Also, default 
normalisation in np.cov is by N-1 (unbiased) wheres in np.std and 
np.var the default is by N (unbiased).

Johann


From mscipy at googlemail.com  Thu Sep 25 08:09:37 2008
From: mscipy at googlemail.com (Saber Mbarek)
Date: Thu, 25 Sep 2008 12:09:37 +0000
Subject: [SciPy-user] importing pylab??
In-Reply-To: <3d375d730809240114m65fc8195n9857ab4c6d590395@mail.gmail.com>
References: <90ec20e0809240057h64f91941t6a246678a098be43@mail.gmail.com>
	<3d375d730809240114m65fc8195n9857ab4c6d590395@mail.gmail.com>
Message-ID: <90ec20e0809250509j5d8e521p9235260f422b0afa@mail.gmail.com>

the import error of the pylab module was related to my locale settings and
not to matplotlib.

thank you for your help
Saber
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080925/439617b8/attachment.html>

From aisaac at american.edu  Thu Sep 25 09:12:56 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Thu, 25 Sep 2008 09:12:56 -0400
Subject: [SciPy-user] Inconsistent standard deviation and
 variance	implementation in scipy vs. scipy.stats
In-Reply-To: <9457e7c80809250048r5bec3240x35d25c8c791ee728@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>	<200809241205.02637.pgmdevlist@gmail.com>	<48DA9519.1000203@american.edu>	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
	<9457e7c80809250048r5bec3240x35d25c8c791ee728@mail.gmail.com>
Message-ID: <48DB8E58.2050304@american.edu>

> 2008/9/24 Anne Archibald <peridot.faceted at gmail.com>:
>> How about (possibly in addition to your suggestion) deprecating the 
>> re-exporting of numpy functions inside scipy? People often seem to ask 
>> about whether they should be using the scipy "version" or the numpy 
>> "version" of some function, when in fact it's just a re-exporting of 
>> the name. 

On 9/25/2008 3:48 AM St?fan van der Walt apparently wrote:
> I'm all in favour of that suggestion. 

I am not taking a strong position other than to say user
convenience should matter, but the following really seems
adequate to me:

     >>> help(sp.var)
     Help on function var in module numpy.core.fromnumeric:

Cheers,
Alan Isaac




From millman at berkeley.edu  Thu Sep 25 15:51:50 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Thu, 25 Sep 2008 12:51:50 -0700
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <9457e7c80809250050u15ca8517o727f87936236da78@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<5b8d13220809242135w7174b0f3l81cf81bddbd3ed89@mail.gmail.com>
	<9457e7c80809250050u15ca8517o727f87936236da78@mail.gmail.com>
Message-ID: <c7009a550809251251n5deb0e6o3e7374dee6887013@mail.gmail.com>

On Thu, Sep 25, 2008 at 12:50 AM, St?fan van der Walt <stefan at sun.ac.za> wrote:
> 2008/9/25 David Cournapeau <cournape at gmail.com>:
>> I suggested some time ago to deprecate scipy.stats current signature
>> for 0.7, and set the new one in 0.8. If that's fine with you, we could
>> do that. I don't feel confortable changing a function in scipy.stats
>> (because it is not "my" module), but OTOH, nobody reacted last time we
>> had this discussion, so maybe we should just do it.
>
> Yes, do it.  If someone complains, that's why we have SVN.  The SciPy
> API is not (and cannot be) frozen -- it is still too immature, so
> let's get it up to scratch ASAP.

I also don't think it is absolutely necessary to have a deprecation
release.  The code is officially labeled beta and is currently not
regularly released.

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


From peridot.faceted at gmail.com  Thu Sep 25 18:30:06 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Thu, 25 Sep 2008 18:30:06 -0400
Subject: [SciPy-user] Documentation
Message-ID: <ce557a360809251530w1f4b20aeode8502b97e5bf877@mail.gmail.com>

Hi,

I've just written a description of how scipy's smoothing splines work
(because a friend is publishing a paper that used them). I had to go
look up the original research papers, because there isn't really a
good description in the python documentation, or even in the FITPACK
source. It seems like it would be valuable to have such a description,
but I'm not sure where it should go: duplicated in every one of
scipy's spline-fitting functions? at the module level? in some
automatic docstring transmogrifier that puts it in all relevant
docstrings without requiring source code duplication?

Anne
"""
The splines constructed by this code are described in a number of
papers (see below). Summarizing based on Dierckx 1982, the goal is to
find a spline for which the "smoothness" (sum of squares of jumps in
the highest derivative at the joins) is as small as possible given
that the (weighted) sum of squares of residuals is S. This can be
achieved by choosing one knot per data point, but in the interests of
efficiency and smoothness, the algorithm attempts to select the
smallest set of knots that can achieve this balance.

For any given set of knots, by adjusting the relative importance of
smoothness and good fitting the algorithm can go from the
least-squares polynomial of degree K (totally smooth) to the
least-squares spline (big jumps at each knot).  The code tries to find
the smallest set of knots for which the least-squares spline gives you
residuals better than S, then adjusts the relative importance of
smoothing and quality-of-fit to make the curve smoother until the
residuals are exactly S.

It finds this best set of knots by starting with the bare minimum of
knots - a one-segment spline - and checking whether the least-squares
spline (with no smoothness constraints) fits the data as well as S. If
not, then it subdivides the interval by introducing a knot. The
subdivided spline is again checked to see whether it can fit the data
with sum of squares of residuals no worse than S. If not, the
worst-fitting subintervals are subdivided again. This is repeated
until there are enough knots to make it possible to fit with residuals
no worse than S. When this is finally achieved - if necessary by
having a knot at every data point so the spline interpolates them -
the spline is adjusted to make it smoother but the fit worse until a
spline is obtained with sum of squares of residuals exactly S. The
procedure does not attempt to produce the strictly minimal set of
knots, but it does stop introducing new knots as soon as possible.

The process is described in detail in:

   Dierckx P. : An Algorithm for Smoothing, Differentiation and Integ-
                ration Of Experimental Data Using Spline Functions,
                J.Comp.Appl.Maths 1 (1975) 165-184.
   Dierckx P. : A Fast Algorithm for Smoothing Data on a Rectangular
                Grid While Using Spline Functions, Siam J.Numer.Anal.
                19 (1982) 1286-1304.
   Dierckx P. : An Improved Algorithm for Curve Fitting with Spline
                Functions, Report TW54, Dept. Computer Science, K.U.
                Leuven, 1981.
   Dierckx P. : Curve and Surface Fitting with Splines, Monographs on
                Numerical Analysis, Oxford University Press, 1993.

The code in scipy is a python wrapper of the FITPACK routines written by:

    P. Dierckx
    Dept. Computer Science, K.U. Leuven
    Celestijnenlaan 200a, B-3001 Heverlee, Belgium.
    e-mail : Paul.Dierckx at cs.kuleuven.ac.be
"""


From robert.kern at gmail.com  Thu Sep 25 18:57:34 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 25 Sep 2008 17:57:34 -0500
Subject: [SciPy-user] Documentation
In-Reply-To: <ce557a360809251530w1f4b20aeode8502b97e5bf877@mail.gmail.com>
References: <ce557a360809251530w1f4b20aeode8502b97e5bf877@mail.gmail.com>
Message-ID: <3d375d730809251557y6ab4aef1vf7bea52db17665cf@mail.gmail.com>

On Thu, Sep 25, 2008 at 17:30, Anne Archibald <peridot.faceted at gmail.com> wrote:
> Hi,
>
> I've just written a description of how scipy's smoothing splines work
> (because a friend is publishing a paper that used them). I had to go
> look up the original research papers, because there isn't really a
> good description in the python documentation, or even in the FITPACK
> source. It seems like it would be valuable to have such a description,
> but I'm not sure where it should go: duplicated in every one of
> scipy's spline-fitting functions? at the module level? in some
> automatic docstring transmogrifier that puts it in all relevant
> docstrings without requiring source code duplication?

We could probably put it in the module docstring and then say "See the
docstring for scipy.interpolate.fitpack2 for details about the
algorithm." in the individual functions.

-- 
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 millman at berkeley.edu  Thu Sep 25 19:01:25 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Thu, 25 Sep 2008 16:01:25 -0700
Subject: [SciPy-user] Documentation
In-Reply-To: <3d375d730809251557y6ab4aef1vf7bea52db17665cf@mail.gmail.com>
References: <ce557a360809251530w1f4b20aeode8502b97e5bf877@mail.gmail.com>
	<3d375d730809251557y6ab4aef1vf7bea52db17665cf@mail.gmail.com>
Message-ID: <c7009a550809251601j20c10922r583a4dd0a1d63436@mail.gmail.com>

On Thu, Sep 25, 2008 at 3:57 PM, Robert Kern <robert.kern at gmail.com> wrote:
> We could probably put it in the module docstring and then say "See the
> docstring for scipy.interpolate.fitpack2 for details about the
> algorithm." in the individual functions.

+1

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


From craigmaloney at cmu.edu  Thu Sep 25 21:46:47 2008
From: craigmaloney at cmu.edu (Craig Maloney)
Date: Thu, 25 Sep 2008 21:46:47 -0400
Subject: [SciPy-user] Tester import errors
Message-ID: <F4143CF7-813E-4257-906C-335B8FA4BA57@cmu.edu>

I'm also getting annoying:

ImportError: cannot import name Tester

Which can be defeated by going into the installed .py files and  
killing the "import Tester" line and subsequent tests.

-----------------------------------------------
Contact info:
http://www.ce.cmu.edu/~maloney2/





From robert.kern at gmail.com  Thu Sep 25 22:01:18 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 25 Sep 2008 21:01:18 -0500
Subject: [SciPy-user] Tester import errors
In-Reply-To: <F4143CF7-813E-4257-906C-335B8FA4BA57@cmu.edu>
References: <F4143CF7-813E-4257-906C-335B8FA4BA57@cmu.edu>
Message-ID: <3d375d730809251901s5cd25021vb64a351a99a5b560@mail.gmail.com>

On Thu, Sep 25, 2008 at 20:46, Craig Maloney <craigmaloney at cmu.edu> wrote:
> I'm also getting annoying:
>
> ImportError: cannot import name Tester
>
> Which can be defeated by going into the installed .py files and
> killing the "import Tester" line and subsequent tests.

Can you provide more context? What versions of numpy and scipy do you
have? What tests are you running?

-- 
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 craigmaloney at cmu.edu  Thu Sep 25 21:44:19 2008
From: craigmaloney at cmu.edu (Craig Maloney)
Date: Thu, 25 Sep 2008 21:44:19 -0400
Subject: [SciPy-user] can't import scipy.sparse.linalg
Message-ID: <6705CF56-540D-419F-9152-DB07B7422893@cmu.edu>

Hi all.

I just built the latest svn version and installed.

I can import, e.g. scipy.sparse.sparsetools, but cannot import  
scipy.sparse.linalg.  No obvious error messages during the build.

Anyone else have this problem?

Looks like there was some reasonably heavy developer activity in this  
directory last week.

(I'm now trying again with svn revision 4736... before last week's  
changes)

Thanks,
Craig

-----------------------------------------------
Contact info:
http://www.ce.cmu.edu/~maloney2/





From wnbell at gmail.com  Thu Sep 25 22:46:00 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Thu, 25 Sep 2008 22:46:00 -0400
Subject: [SciPy-user] can't import scipy.sparse.linalg
In-Reply-To: <6705CF56-540D-419F-9152-DB07B7422893@cmu.edu>
References: <6705CF56-540D-419F-9152-DB07B7422893@cmu.edu>
Message-ID: <d05265cb0809251946y638af15bpd8d4caaf1082deb1@mail.gmail.com>

On Thu, Sep 25, 2008 at 9:44 PM, Craig Maloney <craigmaloney at cmu.edu> wrote:
> Hi all.
>
> I just built the latest svn version and installed.
>
> I can import, e.g. scipy.sparse.sparsetools, but cannot import
> scipy.sparse.linalg.  No obvious error messages during the build.
>
> Anyone else have this problem?
>
> Looks like there was some reasonably heavy developer activity in this
> directory last week.
>
> (I'm now trying again with svn revision 4736... before last week's
> changes)
>

Hi Craig, I just tested r4741 and everything seems to work.  Can you
delete your build directory and site-packages/scipy directory and try
again?


-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From david at ar.media.kyoto-u.ac.jp  Thu Sep 25 23:53:24 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 26 Sep 2008 12:53:24 +0900
Subject: [SciPy-user] Inconsistent standard deviation and
 variance	implementation in scipy vs. scipy.stats
In-Reply-To: <c7009a550809251251n5deb0e6o3e7374dee6887013@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>	<200809241205.02637.pgmdevlist@gmail.com>	<48DA9519.1000203@american.edu>	<5b8d13220809242135w7174b0f3l81cf81bddbd3ed89@mail.gmail.com>	<9457e7c80809250050u15ca8517o727f87936236da78@mail.gmail.com>
	<c7009a550809251251n5deb0e6o3e7374dee6887013@mail.gmail.com>
Message-ID: <48DC5CB4.30202@ar.media.kyoto-u.ac.jp>

Jarrod Millman wrote:
>
> I also don't think it is absolutely necessary to have a deprecation
> release.  The code is officially labeled beta and is currently not
> regularly released.
>   

Yes, but in that case, it would be pretty bad to be caught by it.
Deprecating does not cost us anything (well, it cost me a couple of
minutes), and cost a lot to users. I *hated* it when numpy changed its
axis argument before the 1. release, it took me hours to track it down
everywhere in my code (missing argument is different: you know that
something is wrong right away). I would prefer not inflicting this to
other people. Not in my name, at least :)

cheers,

David



From david at ar.media.kyoto-u.ac.jp  Thu Sep 25 23:55:48 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 26 Sep 2008 12:55:48 +0900
Subject: [SciPy-user] Inconsistent standard deviation and
 variance	implementation in scipy vs. scipy.stats
In-Reply-To: <c7009a550809251251n5deb0e6o3e7374dee6887013@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>	<200809241205.02637.pgmdevlist@gmail.com>	<48DA9519.1000203@american.edu>	<5b8d13220809242135w7174b0f3l81cf81bddbd3ed89@mail.gmail.com>	<9457e7c80809250050u15ca8517o727f87936236da78@mail.gmail.com>
	<c7009a550809251251n5deb0e6o3e7374dee6887013@mail.gmail.com>
Message-ID: <48DC5D44.6000008@ar.media.kyoto-u.ac.jp>

Jarrod Millman wrote:
> On Thu, Sep 25, 2008 at 12:50 AM, St?fan van der Walt <stefan at sun.ac.za> wrote:
>> 2008/9/25 David Cournapeau <cournape at gmail.com>:
>>> I suggested some time ago to deprecate scipy.stats current signature
>>> for 0.7, and set the new one in 0.8. If that's fine with you, we could
>>> do that. I don't feel confortable changing a function in scipy.stats
>>> (because it is not "my" module), but OTOH, nobody reacted last time we
>>> had this discussion, so maybe we should just do it.
>> Yes, do it.  If someone complains, that's why we have SVN.  The SciPy
>> API is not (and cannot be) frozen -- it is still too immature, so
>> let's get it up to scratch ASAP.

I added a Deprecation warning with the correct function to use in numpy
(with alternative arguments) for the following:
    - mean
    - median
    - std
    - var
    - cov
    - corrcoeff
AFAICS, all functionality of any of those is available in numpy
(contrary to what the comment says, I guess they are vastly out of date),

cheers,

David


From cournape at gmail.com  Fri Sep 26 00:19:00 2008
From: cournape at gmail.com (David Cournapeau)
Date: Fri, 26 Sep 2008 13:19:00 +0900
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
Message-ID: <5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>

On Thu, Sep 25, 2008 at 5:35 AM, Anne Archibald
<peridot.faceted at gmail.com> wrote:
> How about (possibly in addition to your suggestion) deprecating the
> re-exporting of numpy functions inside scipy? People often seem to ask
> about whether they should be using the scipy "version" or the numpy
> "version" of some function, when in fact it's just a re-exporting of
> the name.

Yes, it would be nice. What do other people think about deprecating
all the numpy re-export in scipy ? It would be nice to do for 0.7
(e.g. in 0.7, deprecated, in 0.8, removed).

cheers,

David


From jh at physics.ucf.edu  Fri Sep 26 01:15:54 2008
From: jh at physics.ucf.edu (jh at physics.ucf.edu)
Date: Fri, 26 Sep 2008 01:15:54 -0400
Subject: [SciPy-user] Documentation (Anne Archibald)
In-Reply-To: <mailman.137.1222402744.2878.scipy-user@scipy.org>
	(scipy-user-request@scipy.org)
References: <mailman.137.1222402744.2878.scipy-user@scipy.org>
Message-ID: <wl8d4irlc6t.fsf@glup.Belkin>

Anne writes:

> It seems like it would be valuable to have such a description,
> but I'm not sure where it should go: duplicated in every one of
> scipy's spline-fitting functions? at the module level? in some
> automatic docstring transmogrifier that puts it in all relevant
> docstrings without requiring source code duplication?

First, thanks for researching and writing this!

This is going to become a general issue, particularly for scipy, where
there may be need for many of these in certain modules.  What we
choose for this one should scale to having many of these per
sub-module in some cases, and just one or none in others.  I would
think that the first one or two per sub-module could go in the
sub-module's docstring as you suggest, but after that we should
separate all of them into either a .doc sub-sub-module for that
sub-module (as now exists for numpy itself) or a function called
doc_splines (in this case) that contains only the docstring and
"pass".  Either way, those pages should be referenced in the
sub-module docstring so people know they exist.

If we do the doc-function, which is admittedly a hack but which may be
more convenient than having to do yet another import, the one-line
description should say something appropriate to identify it in lists
of functions, something like "documentation for spline
implementation".  If we could make np.info() import anything it needs
to in order to access the named docstring, it would make life easier
and the .doc sub-sub-module more attractive, since you wouldn't have
to import it.  In this case, np.info(sp.foo.doc.spline) would import
sp.foo.doc automatically.

Whatever is decided, a description should go in the doc standards.

--jh--


From s.mientki at ru.nl  Fri Sep 26 04:09:08 2008
From: s.mientki at ru.nl (Stef Mientki)
Date: Fri, 26 Sep 2008 10:09:08 +0200
Subject: [SciPy-user] Documentation
In-Reply-To: <ce557a360809251530w1f4b20aeode8502b97e5bf877@mail.gmail.com>
References: <ce557a360809251530w1f4b20aeode8502b97e5bf877@mail.gmail.com>
Message-ID: <48DC98A4.6030805@ru.nl>



Anne Archibald wrote:
> Hi,
>
> I've just written a description of how scipy's smoothing splines work
> (because a friend is publishing a paper that used them). I had to go
> look up the original research papers, because there isn't really a
> good description in the python documentation, or even in the FITPACK
> source. It seems like it would be valuable to have such a description,
> but I'm not sure where it should go: duplicated in every one of
> scipy's spline-fitting functions? at the module level? in some
> automatic docstring transmogrifier that puts it in all relevant
> docstrings without requiring source code duplication?
>
> Anne
>   

Somewhat more documentation would be welcome.

But should we limit ourself to these doc-strings,
which will undoubtedly be a huge amount of work,
even to get a small part of Scipy documented.

Coming from MatLab, I always found the help (if you know the right 
keywords) very good,
it not only described the use of a function, but also fave a lot of 
information about the background.

As I'm in the middle of creating a Matlab like environment in Python,
and the following ideas are now playing through my head:
- the help function shows of course the docstring
- the users personal notes about this function (probably stored in html 
or simple-rtf)
- if the help for this function is asked for the first time, the program 
will search the web.
The websearch is done on a number of sites defined by the user,
and each first search the user can select a number of sites from this list.
For the current subject and each selected website he can also specify of 
the search has to be performed each time.
An example of these sites:
= wikipedia
= some Scipy site (Enthought?)
= google
= MatWorks / Wolfram or is this unethical ?
The found links are opened in a tabbed browser and also stored  in the 
user personal notes.
- the found pages can also be automatically copied to the users own 
domain (how many links sti;; work after a year ?)

So users collects slowly all the information he's interested in,
and can organize the information in his own way.

If I dream further,
the information collected by each user is automatically gathered 
somewhere on the web,
organized by it's popularity,
and thereby forming a good starting point for each newbie.

cheers,
Stef







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 williams at astro.ox.ac.uk  Fri Sep 26 18:55:35 2008
From: williams at astro.ox.ac.uk (Michael Williams)
Date: Fri, 26 Sep 2008 23:55:35 +0100
Subject: [SciPy-user] Installation from source on OS X:
	'NoneType'	object has no attribute 'link_shared_object'
In-Reply-To: <3d375d730809231153p43290a4bu10cdf119908c8830@mail.gmail.com>
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
	<3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
	<BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>
	<3d375d730809231153p43290a4bu10cdf119908c8830@mail.gmail.com>
Message-ID: <20080926225534.GA6402@astro.ox.ac.uk>

On Tue, Sep 23, 2008 at 01:53:21PM -0500, Robert Kern wrote:
> Unfortunately, now it looks like you are dealing with a bug in
> gfortran. Can you downgrade to something earlier than 4.3.0?

With difficulty (hence the delayed reply). I'm trying to install without
user privileges and I'm not experienced with gcc installation.

That being said, I seem to have succeeded. I installed gfortran 4.4.0
from the HPC on OS X page into ~/usr/local (rather than /usr/local),
added it to my shell's path and ran setup.py with the usual gnu95
options. It seemed to build and install fine. (Does anyone less ignorant
than I am about compilers and libraries know if I need to keep the
contents of ~/usr/local around to use scipy?)

I can import cleanly (Deprecation Warning about NumpyTest
notwithstanding). scipy.test runs 1848 tests with one failure
(check_dot). Last time I installed scipy I think most people were not
too worried by errors in scipy.test() on OS X. Is that still the case?

Thanks again, 

-- Mike


From robert.kern at gmail.com  Fri Sep 26 19:14:21 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 26 Sep 2008 18:14:21 -0500
Subject: [SciPy-user] Installation from source on OS X: 'NoneType'
	object has no attribute 'link_shared_object'
In-Reply-To: <20080926225534.GA6402@astro.ox.ac.uk>
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
	<3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
	<BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>
	<3d375d730809231153p43290a4bu10cdf119908c8830@mail.gmail.com>
	<20080926225534.GA6402@astro.ox.ac.uk>
Message-ID: <3d375d730809261614j7696ced5je18889036ea67fa1@mail.gmail.com>

On Fri, Sep 26, 2008 at 17:55, Michael Williams <williams at astro.ox.ac.uk> wrote:
> On Tue, Sep 23, 2008 at 01:53:21PM -0500, Robert Kern wrote:
>> Unfortunately, now it looks like you are dealing with a bug in
>> gfortran. Can you downgrade to something earlier than 4.3.0?
>
> With difficulty (hence the delayed reply). I'm trying to install without
> user privileges and I'm not experienced with gcc installation.
>
> That being said, I seem to have succeeded. I installed gfortran 4.4.0
> from the HPC on OS X page into ~/usr/local (rather than /usr/local),
> added it to my shell's path and ran setup.py with the usual gnu95
> options. It seemed to build and install fine. (Does anyone less ignorant
> than I am about compilers and libraries know if I need to keep the
> contents of ~/usr/local around to use scipy?)

Yes. gfortran will link against the shared libraries there unless if
you do a fairly complicated dance to make it link against the static
libraries.

I very much recommend against using the binaries from HPC. They
release binaries for buggy bleeding-edge versions of gfortran, and
don't keep previous versions around. I have had *much* more luck with
the binaries over here:

  http://r.research.att.com/tools/

Since you don't have admin privileges, you need to do a little command
line work instead of just being able to use the installer. Mount
gfortran-4.2.3.dmg. At the terminal:

  # Assuming you want it in ~/usr/local/
  $ cd ~/
  $ pax -zr < /Volumes/GNU\ Fortran\ 4.2.3/gfortran.pkg/Contents/Archive.pax.gz

Now you should have everything unpacked into ~/usr/local/.

But if everything is currently working for you, I would probably leave
well enough alone.

> I can import cleanly (Deprecation Warning about NumpyTest
> notwithstanding). scipy.test runs 1848 tests with one failure
> (check_dot). Last time I installed scipy I think most people were not
> too worried by errors in scipy.test() on OS X. Is that still the case?

You're probably not using that function, so it's fine in that sense.
But it is a bona fide bug in 0.6.0 and has been fixed on the trunk.

-- 
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 williams at astro.ox.ac.uk  Fri Sep 26 20:02:37 2008
From: williams at astro.ox.ac.uk (Michael Williams)
Date: Sat, 27 Sep 2008 01:02:37 +0100
Subject: [SciPy-user] Installation from source on OS X: 'NoneType'
	object has no attribute 'link_shared_object'
In-Reply-To: <3d375d730809261614j7696ced5je18889036ea67fa1@mail.gmail.com>
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
	<3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
	<BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>
	<3d375d730809231153p43290a4bu10cdf119908c8830@mail.gmail.com>
	<20080926225534.GA6402@astro.ox.ac.uk>
	<3d375d730809261614j7696ced5je18889036ea67fa1@mail.gmail.com>
Message-ID: <5B3D66A5-B141-4DE7-99B3-A53F5EEE00E8@astro.ox.ac.uk>


On 27 Sep 2008, at 00:14, Robert Kern wrote:
> I very much recommend against using the binaries from HPC. They
> release binaries for buggy bleeding-edge versions of gfortran, and
> don't keep previous versions around.

Thanks for the heads-up. I'm not experienced with Fortran.

> I have had *much* more luck with
> the binaries over here:
>
>  http://r.research.att.com/tools/
>
> Since you don't have admin privileges, you need to do a little command
> line work instead of just being able to use the installer. Mount
> gfortran-4.2.3.dmg. At the terminal:

That procedure to install the compiler worked fine, and scipy itself  
seems to have built cleanly using it.

Thanks very much!

-- Mike




From matthias.blaicher at googlemail.com  Sat Sep 27 10:58:51 2008
From: matthias.blaicher at googlemail.com (Matthias Blaicher)
Date: Sat, 27 Sep 2008 16:58:51 +0200
Subject: [SciPy-user] scipy/numpy on CentOS 5/ Dependency problem
Message-ID: <261cc8ff0809270758p30f1f259r2481497402539c2d@mail.gmail.com>

Hello,

I want to install SciPy and numpy on a CentOS 5 system. It contains a
fresh install with up-to-date packages. I use the "official" Suse
Build repos by DavidCournapeau.

It fails with a conflict between refblas3 and blas.

[root at localhost yum.repos.d]# yum install python-numpy python-scipy
Loading "fastestmirror" plugin
Loading mirror speeds from cached hostfile
 * home_ashigabou: download.opensuse.org
 * extras: mirrors.tummy.com
 * updates: dds.gina.alaska.edu
 * base: mirror.hmc.edu
 * addons: mirrors.tummy.com
extras                    100% |=========================| 1.1 kB    00:00
updates                   100% |=========================|  951 B    00:00
base                      100% |=========================| 1.1 kB    00:00
addons                    100% |=========================|  951 B    00:00
Setting up Install Process
Parsing package install arguments
Resolving Dependencies
--> Running transaction check
---> Package python-scipy.i386 0:0.6.0-2.1 set to be updated
--> Processing Dependency: libblas.so.3 for package: python-scipy
--> Processing Dependency: libblas.so.3 for package: python-scipy
--> Processing Dependency: liblapack.so.3 for package: python-scipy
--> Processing Dependency: libgfortran.so.1 for package: python-scipy
--> Processing Dependency: liblapack.so.3 for package: python-scipy
---> Package python-numpy.i386 0:1.2.0-1.1 set to be updated
--> Processing Dependency: gcc-gfortran for package: python-numpy
--> Processing Dependency: refblas3 for package: python-numpy
--> Processing Dependency: lapack3 < 3.1 for package: python-numpy
--> Running transaction check
---> Package gcc-gfortran.i386 0:4.1.2-42.el5 set to be updated
--> Processing Dependency: gcc = 4.1.2-42.el5 for package: gcc-gfortran
---> Package blas.i386 0:3.0-37.el5 set to be updated
---> Package lapack3.i386 0:3.0-19.1 set to be updated
---> Package lapack.i386 0:3.0-37.el5 set to be updated
---> Package refblas3.i386 0:3.0-11.1 set to be updated
---> Package libgfortran.i386 0:4.1.2-42.el5 set to be updated
--> Running transaction check
---> Package gcc.i386 0:4.1.2-42.el5 set to be updated
--> Processing Dependency: libgomp.so.1 for package: gcc
--> Processing Dependency: glibc-devel >= 2.2.90-12 for package: gcc
--> Processing Dependency: libgomp = 4.1.2-42.el5 for package: gcc
--> Running transaction check
---> Package glibc-devel.i386 0:2.5-24 set to be updated
--> Processing Dependency: glibc-headers for package: glibc-devel
--> Processing Dependency: glibc-headers = 2.5-24 for package: glibc-devel
---> Package libgomp.i386 0:4.1.2-42.el5 set to be updated
--> Running transaction check
---> Package glibc-headers.i386 0:2.5-24 set to be updated
--> Processing Dependency: kernel-headers for package: glibc-headers
--> Processing Dependency: kernel-headers >= 2.2.1 for package: glibc-headers
--> Running transaction check
---> Package kernel-headers.i386 0:2.6.18-92.1.13.el5 set to be updated
--> Processing Conflict: refblas3 conflicts blas
--> Finished Dependency Resolution
Error: refblas3 conflicts with blas

Is there anything I'm doing wrong here, as I don't want to compile all
dependencies by myself..

Sincerly,

Matthias Blaicher


From matthieu.brucher at gmail.com  Sat Sep 27 11:03:55 2008
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Sat, 27 Sep 2008 17:03:55 +0200
Subject: [SciPy-user] scipy/numpy on CentOS 5/ Dependency problem
In-Reply-To: <261cc8ff0809270758p30f1f259r2481497402539c2d@mail.gmail.com>
References: <261cc8ff0809270758p30f1f259r2481497402539c2d@mail.gmail.com>
Message-ID: <e76aa17f0809270803n41d2bfdep8b29e9dfd1a4a022@mail.gmail.com>

Hi,

Start by installing refblas3, refblas3-dev, lapack3 and lapack3-dev.
Last time I had this issue, installing the dependencies by hand first
solved it.

Matthieu

2008/9/27 Matthias Blaicher <matthias.blaicher at googlemail.com>:
> Hello,
>
> I want to install SciPy and numpy on a CentOS 5 system. It contains a
> fresh install with up-to-date packages. I use the "official" Suse
> Build repos by DavidCournapeau.
>
> It fails with a conflict between refblas3 and blas.
>
> [root at localhost yum.repos.d]# yum install python-numpy python-scipy
> Loading "fastestmirror" plugin
> Loading mirror speeds from cached hostfile
>  * home_ashigabou: download.opensuse.org
>  * extras: mirrors.tummy.com
>  * updates: dds.gina.alaska.edu
>  * base: mirror.hmc.edu
>  * addons: mirrors.tummy.com
> extras                    100% |=========================| 1.1 kB    00:00
> updates                   100% |=========================|  951 B    00:00
> base                      100% |=========================| 1.1 kB    00:00
> addons                    100% |=========================|  951 B    00:00
> Setting up Install Process
> Parsing package install arguments
> Resolving Dependencies
> --> Running transaction check
> ---> Package python-scipy.i386 0:0.6.0-2.1 set to be updated
> --> Processing Dependency: libblas.so.3 for package: python-scipy
> --> Processing Dependency: libblas.so.3 for package: python-scipy
> --> Processing Dependency: liblapack.so.3 for package: python-scipy
> --> Processing Dependency: libgfortran.so.1 for package: python-scipy
> --> Processing Dependency: liblapack.so.3 for package: python-scipy
> ---> Package python-numpy.i386 0:1.2.0-1.1 set to be updated
> --> Processing Dependency: gcc-gfortran for package: python-numpy
> --> Processing Dependency: refblas3 for package: python-numpy
> --> Processing Dependency: lapack3 < 3.1 for package: python-numpy
> --> Running transaction check
> ---> Package gcc-gfortran.i386 0:4.1.2-42.el5 set to be updated
> --> Processing Dependency: gcc = 4.1.2-42.el5 for package: gcc-gfortran
> ---> Package blas.i386 0:3.0-37.el5 set to be updated
> ---> Package lapack3.i386 0:3.0-19.1 set to be updated
> ---> Package lapack.i386 0:3.0-37.el5 set to be updated
> ---> Package refblas3.i386 0:3.0-11.1 set to be updated
> ---> Package libgfortran.i386 0:4.1.2-42.el5 set to be updated
> --> Running transaction check
> ---> Package gcc.i386 0:4.1.2-42.el5 set to be updated
> --> Processing Dependency: libgomp.so.1 for package: gcc
> --> Processing Dependency: glibc-devel >= 2.2.90-12 for package: gcc
> --> Processing Dependency: libgomp = 4.1.2-42.el5 for package: gcc
> --> Running transaction check
> ---> Package glibc-devel.i386 0:2.5-24 set to be updated
> --> Processing Dependency: glibc-headers for package: glibc-devel
> --> Processing Dependency: glibc-headers = 2.5-24 for package: glibc-devel
> ---> Package libgomp.i386 0:4.1.2-42.el5 set to be updated
> --> Running transaction check
> ---> Package glibc-headers.i386 0:2.5-24 set to be updated
> --> Processing Dependency: kernel-headers for package: glibc-headers
> --> Processing Dependency: kernel-headers >= 2.2.1 for package: glibc-headers
> --> Running transaction check
> ---> Package kernel-headers.i386 0:2.6.18-92.1.13.el5 set to be updated
> --> Processing Conflict: refblas3 conflicts blas
> --> Finished Dependency Resolution
> Error: refblas3 conflicts with blas
>
> Is there anything I'm doing wrong here, as I don't want to compile all
> dependencies by myself..
>
> Sincerly,
>
> Matthias Blaicher
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



-- 
French PhD student
Information System Engineer
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 contact at pythonxy.com  Sat Sep 27 12:09:20 2008
From: contact at pythonxy.com (Pierre Raybaut)
Date: Sat, 27 Sep 2008 18:09:20 +0200
Subject: [SciPy-user] [ Python(x,y) ] Next release
Message-ID: <48DE5AB0.7000106@pythonxy.com>

Hi all,

As you may already know, Python(x,y) is a free scientific-oriented
Python Distribution based on Qt and Eclipse providing a self-consistent
scientific development environment.

Release 2.1.0 has introduced a new plugin system allowing easy update 
(http://www.pythonxy.com/standard.php) and fully customizable installation:
http://www.pythonxy.com/_tools/img.php?lang=1&img=/_images/Windows%20Installer/03b.png
http://www.pythonxy.com/screenshots.php?dir=/Windows%20Installer

Third-party plugins are now available 
(http://www.pythonxy.com/additional.php) - please do not hesitate to 
contribute and create your own thanks to the SDK.

 From now on, releases will be scheduled monthly.
However, next release will somehow be available continously because 
future plugins will be available separately. For example, NumPy 1.2.0, 
VTK 5.2.0 or ETS 3.0.2 are already available at 
http://www.pythonxy.com/additional.php.

Regards,
Pierre Raybaut



From contact at pythonxy.com  Sun Sep 28 14:11:23 2008
From: contact at pythonxy.com (Pierre Raybaut)
Date: Sun, 28 Sep 2008 20:11:23 +0200
Subject: [SciPy-user] [ Python(x,y) ] New release : 2.1.1
Message-ID: <48DFC8CB.8080009@pythonxy.com>

Hi all,

As you may already know, Python(x,y) is a free scientific-oriented
Python Distribution based on Qt and Eclipse providing a self-consistent
scientific development environment.

Release 2.1.1 is now available on http://www.pythonxy.com.
(Full Edition, Basic Edition, Light Edition, Custom Edition and Update)

Changes history
Version 2.1.1 (09-28-2008)
     * Added:
           o Sphinx 0.4.2
           o nose 0.10.3
     * Updated:
           o NumPy 1.2.0
           o VTK 5.2.0
           o Enthought Tool Suite 3.0.2
           o GDCM 2.0.9
           o setuptools 0.6c9
           o xy 1.0.6.1
           o IPython 0.9.1.1
           o Console 2.0.140.4
     * Corrected:
           o Issues 18, 19, 20 (see 'Issues' section on Python(x,y) 
Google Code website)

Regards,
Pierre Raybaut



From gary.pajer at gmail.com  Sun Sep 28 15:05:12 2008
From: gary.pajer at gmail.com (Gary Pajer)
Date: Sun, 28 Sep 2008 15:05:12 -0400
Subject: [SciPy-user] NumpyTest warning after upgrade
Message-ID: <88fe22a0809281205p11155ff8l2204ffa5d63011ff@mail.gmail.com>

I've upgraded numpy and scipy after quite some time.  I think I had numpy
1.0.4 and scipy 5.2, but I'm not 100% sure.
I'm on winxp, python 2.5.1, and I installed using the windows exe installers
on scipy.org

I've poked around the scipy-users and numpy-users lists, and learned what I
already know:  NumpyTest is deprecated.  How can I fix this?

regards,
Gary
---------------------------------------------------------------------------------------------------------
C:\Documents and Settings\Gary\My Documents>ipython
Python 2.5.1 (r251:54863, Apr 18 2007, 08:51:08) [MSC v.1310 32 bit (Intel)]
Type "copyright", "credits" or "license" for more information.

IPython 0.9.1 -- An enhanced Interactive Python.
?         -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help      -> Python's own help system.
object?   -> Details about 'object'. ?object also works, ?? prints more.

In [1]: import numpy

In [2]: import scipy
c:\python25\lib\site-packages\scipy\misc\__init__.py:25: DeprecationWarning:
NumpyTest will be removed in the next release; please upda
te your code to use nose or unittest
  test = NumpyTest().test

In [3]: numpy.__version__
Out[3]: '1.2.0'

In [4]: scipy.__version__
Out[4]: '0.6.0'
-------------------------------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080928/15dc1e4d/attachment.html>

From alan.mcintyre at gmail.com  Sun Sep 28 15:36:50 2008
From: alan.mcintyre at gmail.com (Alan McIntyre)
Date: Sun, 28 Sep 2008 12:36:50 -0700
Subject: [SciPy-user] NumpyTest warning after upgrade
In-Reply-To: <88fe22a0809281205p11155ff8l2204ffa5d63011ff@mail.gmail.com>
References: <88fe22a0809281205p11155ff8l2204ffa5d63011ff@mail.gmail.com>
Message-ID: <1d36917a0809281236r5042b66fxddee186e7ec405d2@mail.gmail.com>

On Sun, Sep 28, 2008 at 12:05 PM, Gary Pajer <gary.pajer at gmail.com> wrote:
> I've upgraded numpy and scipy after quite some time.  I think I had numpy
> 1.0.4 and scipy 5.2, but I'm not 100% sure.
> I'm on winxp, python 2.5.1, and I installed using the windows exe installers
> on scipy.org
>
> I've poked around the scipy-users and numpy-users lists, and learned what I
> already know:  NumpyTest is deprecated.  How can I fix this?
>
> regards,
> Gary

Hi Gary,

Some changes were made to the testing framework in NumPy 1.2, but
SciPy 0.6 doesn't include the changes needed to work with NumPy 1.2
without raising some deprecation warnings (I think this my
understanding of the versions here is correct, somebody please correct
me if I'm wrong).   As far as I know, SciPy 0.6 should still work, but
it will just complain a bit. :)  I'm not sure what the timeline is for
releasing the next SciPy.

In the meantime, if you just want to patch up your local install, you
should be able to change code that looks like this in
scipy/misc/__init__.py (and probably several other places):

from numpy.testing import NumpyTest
test = NumpyTest().test

to this:

from numpy.testing import Tester
test = Tester().test

I suppose you could switch to NumPy 1.1 or try using the SciPy from
svn if you'd like to avoid making local tweaks.

Hope this helps,
Alan


From gary.pajer at gmail.com  Sun Sep 28 18:01:34 2008
From: gary.pajer at gmail.com (Gary Pajer)
Date: Sun, 28 Sep 2008 18:01:34 -0400
Subject: [SciPy-user] NumpyTest warning after upgrade
In-Reply-To: <1d36917a0809281236r5042b66fxddee186e7ec405d2@mail.gmail.com>
References: <88fe22a0809281205p11155ff8l2204ffa5d63011ff@mail.gmail.com>
	<1d36917a0809281236r5042b66fxddee186e7ec405d2@mail.gmail.com>
Message-ID: <88fe22a0809281501mf6f5891s604e9e21f6d637ca@mail.gmail.com>

On Sun, Sep 28, 2008 at 3:36 PM, Alan McIntyre <alan.mcintyre at gmail.com>wrote:

> On Sun, Sep 28, 2008 at 12:05 PM, Gary Pajer <gary.pajer at gmail.com> wrote:
> > I've upgraded numpy and scipy after quite some time.  I think I had numpy
> > 1.0.4 and scipy 5.2, but I'm not 100% sure.
> > I'm on winxp, python 2.5.1, and I installed using the windows exe
> installers
> > on scipy.org
> >
> > I've poked around the scipy-users and numpy-users lists, and learned what
> I
> > already know:  NumpyTest is deprecated.  How can I fix this?
> >
> > regards,
> > Gary
>
> Hi Gary,
>
> Some changes were made to the testing framework in NumPy 1.2, but
> SciPy 0.6 doesn't include the changes needed to work with NumPy 1.2
> without raising some deprecation warnings (I think this my
> understanding of the versions here is correct, somebody please correct
> me if I'm wrong).   As far as I know, SciPy 0.6 should still work, but
> it will just complain a bit. :)  I'm not sure what the timeline is for
> releasing the next SciPy.
>
> In the meantime, if you just want to patch up your local install, you
> should be able to change code that looks like this in
> scipy/misc/__init__.py (and probably several other places):
>
> from numpy.testing import NumpyTest
> test = NumpyTest().test
>
> to this:
>
> from numpy.testing import Tester
> test = Tester().test
>
> I suppose you could switch to NumPy 1.1 or try using the SciPy from
> svn if you'd like to avoid making local tweaks.
>
> Hope this helps,
> Alan


OK ... now I see that numpy is new, and scipy hasn't caught up.  I know this
happens from time to time. I haven't been paying attention.

Thanks.


>
> _______________________________________________
> 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/20080928/d3e9d555/attachment.html>

From didier.rano at gmail.com  Sun Sep 28 18:23:49 2008
From: didier.rano at gmail.com (didier rano)
Date: Sun, 28 Sep 2008 18:23:49 -0400
Subject: [SciPy-user] Import problem numpy.ma
Message-ID: <da09baf80809281523x9f5aa76ybdd021e1d9ab9cee@mail.gmail.com>

Hi all,
I am using numpy in trunk subversion (revision 5872). I have an import
problem with "import numpy.ma" => "ImportError: No module named ma"

I am using macos X 10.5.

I don't know dependences with this module.

Thanks for your help

Didier Rano
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080928/50770963/attachment.html>

From joao.q.fonseca at gmail.com  Mon Sep 29 11:31:46 2008
From: joao.q.fonseca at gmail.com (=?ISO-8859-1?Q?Jo=E3o_Quinta_da_Fonseca?=)
Date: Mon, 29 Sep 2008 16:31:46 +0100
Subject: [SciPy-user] scipy workflow
Message-ID: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>

I have given up Matlab and I it feels great to use scipy for my  
scientific computing needs. However, I find that I use scipy just like  
I used matlab, with ipython as my terminal. Although this works OK,  
some things are a little frustrating: updated modules need to be  
reimported after modification, running scripts requires the run  
command etc., but this is to be expected because scipy is not matlbab.
Now I don't want the matlab way necessarily and I am happy to learn a  
new way to do things. I guess what I would like to do is do things the  
python way, which I think means writing my code as classes with the  
_main_ bit at the end etc., but I am not sure whether that is the best  
way and if so, what is the best way to learn it.
Does any one feel like sharing what their workflow with scipy is? Any  
tips for me?

Thanks,

Joao




From pav at iki.fi  Mon Sep 29 11:42:11 2008
From: pav at iki.fi (Pauli Virtanen)
Date: Mon, 29 Sep 2008 15:42:11 +0000 (UTC)
Subject: [SciPy-user] scipy workflow
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
Message-ID: <gbqt0j$45g$1@ger.gmane.org>

Mon, 29 Sep 2008 16:31:46 +0100, Jo?o Quinta da Fonseca wrote:
> I have given up Matlab and I it feels great to use scipy for my
> scientific computing needs. However, I find that I use scipy just like I
> used matlab, with ipython as my terminal. Although this works OK, some
> things are a little frustrating: updated modules need to be reimported
> after modification, running scripts requires the run command etc., but
> this is to be expected because scipy is not matlbab. Now I don't want
> the matlab way necessarily and I am happy to learn a new way to do
> things. I guess what I would like to do is do things the python way,
> which I think means writing my code as classes with the _main_ bit at
> the end etc., but I am not sure whether that is the best way and if so,
> what is the best way to learn it. Does any one feel like sharing what
> their workflow with scipy is? Any tips for me?

My workflow consists of two parts:

1) Long-running scripts
2) Interactive use

For 1), I typically write programs that have main() and which take 
parameters on the command line. No code is at the module level, 
everything is in functions or classes.

For 2), I use Ipython with ipy_autoreload and matplotlib, which gets 
quite close to Matlab-like usage. I can easily import routines from the 
script files when needed, and the autoreload takes care of reloading them 
for me. The usual caveats for reloading modules apply, but most of the 
time they are not a problem.

-- 
Pauli Virtanen



From massimo.sandal at unibo.it  Mon Sep 29 12:07:12 2008
From: massimo.sandal at unibo.it (massimo sandal)
Date: Mon, 29 Sep 2008 18:07:12 +0200
Subject: [SciPy-user] scipy workflow
In-Reply-To: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
Message-ID: <48E0FD30.1020505@unibo.it>

Jo?o Quinta da Fonseca wrote:
> I have given up Matlab and I it feels great to use scipy for my  
> scientific computing needs. However, I find that I use scipy just like  
> I used matlab, with ipython as my terminal. Although this works OK,  
> some things are a little frustrating: updated modules need to be  
> reimported after modification, running scripts requires the run  
> command etc., but this is to be expected because scipy is not matlbab.
> Now I don't want the matlab way necessarily and I am happy to learn a  
> new way to do things. I guess what I would like to do is do things the  
> python way, which I think means writing my code as classes with the  
> _main_ bit at the end etc., but I am not sure whether that is the best  
> way and if so, what is the best way to learn it.
> Does any one feel like sharing what their workflow with scipy is? Any  
> tips for me?

Apart from the data analysis software where I use scipy as a library, I 
practically only write scripts, without interactive use.

I feel there is more control this way. I don't see much purpose in 
interactive use.

However, you can also write procedural scripts (without classes etc.) in 
Python... no need to delve into OO programming if you don't need to 
(although I like OO personally).

m.

-- 
Massimo Sandal , Ph.D.
University of Bologna
Department of Biochemistry "G.Moruzzi"

snail mail:
Via Irnerio 48, 40126 Bologna, Italy

email:
massimo.sandal at unibo.it

web:
http://www.biocfarm.unibo.it/samori/people/sandal.html

tel: +39-051-2094388
fax: +39-051-2094387
-------------- next part --------------
A non-text attachment was scrubbed...
Name: massimo_sandal.vcf
Type: text/x-vcard
Size: 274 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080929/1f06d264/attachment.vcf>

From mhearne at usgs.gov  Mon Sep 29 12:14:01 2008
From: mhearne at usgs.gov (Michael Hearne)
Date: Mon, 29 Sep 2008 10:14:01 -0600
Subject: [SciPy-user] scipy workflow
In-Reply-To: <48E0FD30.1020505@unibo.it>
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
	<48E0FD30.1020505@unibo.it>
Message-ID: <48E0FEC9.8000209@usgs.gov>

My main two uses of ipython are for:
1) Testing code at the command line that I'm trying to write in a 
script.  Regular expressions, for example, are not something I can code 
in my head.
2) Using the debugger (run -d script.py).  This is a huge help for me.  
The debugger interaction is not as good as Matlab's - for example, if 
you want to set a breakpoint in a module contained in a different file 
than your "main", there's a lot of typing.  It would eventually be nice 
to have a graphical debugger, where you could click on the line of code 
to set a breakpoint.

I use "reset" in ipython to re-import modules after modifying the code.  
The only annoying thing about that is getting prompted (yes, I really 
want to blow away all of my variables!)



massimo sandal wrote:
> Jo?o Quinta da Fonseca wrote:
>> I have given up Matlab and I it feels great to use scipy for my  
>> scientific computing needs. However, I find that I use scipy just 
>> like  I used matlab, with ipython as my terminal. Although this works 
>> OK,  some things are a little frustrating: updated modules need to 
>> be  reimported after modification, running scripts requires the run  
>> command etc., but this is to be expected because scipy is not matlbab.
>> Now I don't want the matlab way necessarily and I am happy to learn 
>> a  new way to do things. I guess what I would like to do is do things 
>> the  python way, which I think means writing my code as classes with 
>> the  _main_ bit at the end etc., but I am not sure whether that is 
>> the best  way and if so, what is the best way to learn it.
>> Does any one feel like sharing what their workflow with scipy is? 
>> Any  tips for me?
>
> Apart from the data analysis software where I use scipy as a library, 
> I practically only write scripts, without interactive use.
>
> I feel there is more control this way. I don't see much purpose in 
> interactive use.
>
> However, you can also write procedural scripts (without classes etc.) 
> in Python... no need to delve into OO programming if you don't need to 
> (although I like OO personally).
>
> m.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   

-- 
------------------------------------------------------
Michael Hearne
mhearne at usgs.gov
(303) 273-8620
USGS National Earthquake Information Center
1711 Illinois St. Golden CO 80401
Senior Software Engineer
Synergetics, Inc.
------------------------------------------------------



From barrywark at gmail.com  Mon Sep 29 18:37:55 2008
From: barrywark at gmail.com (Barry Wark)
Date: Mon, 29 Sep 2008 15:37:55 -0700
Subject: [SciPy-user] scipy workflow
In-Reply-To: <48E0FEC9.8000209@usgs.gov>
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
	<48E0FD30.1020505@unibo.it> <48E0FEC9.8000209@usgs.gov>
Message-ID: <cd7634ce0809291537p2991976dxe989401c73121208@mail.gmail.com>

On Mon, Sep 29, 2008 at 9:14 AM, Michael Hearne <mhearne at usgs.gov> wrote:
> My main two uses of ipython are for:
> 1) Testing code at the command line that I'm trying to write in a
> script.  Regular expressions, for example, are not something I can code
> in my head.
> 2) Using the debugger (run -d script.py).  This is a huge help for me.
> The debugger interaction is not as good as Matlab's - for example, if
> you want to set a breakpoint in a module contained in a different file
> than your "main", there's a lot of typing.  It would eventually be nice
> to have a graphical debugger, where you could click on the line of code
> to set a breakpoint.

If you are an emacs user, I urge you to check out the IPython (minor?)
mode, ipython.el (in the ipython distribution) . It allows much
tighter integration of editor and IPython session, more like Matlab.
If you're a Mac user, I'd also urge you to check out TextMate. You can
achieve similar editor/command-line integration in TextMate using
AppleScript (email me offline if you're interested...I haven't gotten
all of the relevant code posted anywhere yet).

-Barry

>
> I use "reset" in ipython to re-import modules after modifying the code.
> The only annoying thing about that is getting prompted (yes, I really
> want to blow away all of my variables!)
>
>
>
> massimo sandal wrote:
>> Jo?o Quinta da Fonseca wrote:
>>> I have given up Matlab and I it feels great to use scipy for my
>>> scientific computing needs. However, I find that I use scipy just
>>> like  I used matlab, with ipython as my terminal. Although this works
>>> OK,  some things are a little frustrating: updated modules need to
>>> be  reimported after modification, running scripts requires the run
>>> command etc., but this is to be expected because scipy is not matlbab.
>>> Now I don't want the matlab way necessarily and I am happy to learn
>>> a  new way to do things. I guess what I would like to do is do things
>>> the  python way, which I think means writing my code as classes with
>>> the  _main_ bit at the end etc., but I am not sure whether that is
>>> the best  way and if so, what is the best way to learn it.
>>> Does any one feel like sharing what their workflow with scipy is?
>>> Any  tips for me?
>>
>> Apart from the data analysis software where I use scipy as a library,
>> I practically only write scripts, without interactive use.
>>
>> I feel there is more control this way. I don't see much purpose in
>> interactive use.
>>
>> However, you can also write procedural scripts (without classes etc.)
>> in Python... no need to delve into OO programming if you don't need to
>> (although I like OO personally).
>>
>> m.
>>
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>
> --
> ------------------------------------------------------
> Michael Hearne
> mhearne at usgs.gov
> (303) 273-8620
> USGS National Earthquake Information Center
> 1711 Illinois St. Golden CO 80401
> Senior Software Engineer
> Synergetics, Inc.
> ------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From s.mientki at ru.nl  Mon Sep 29 17:09:21 2008
From: s.mientki at ru.nl (Stef Mientki)
Date: Mon, 29 Sep 2008 23:09:21 +0200
Subject: [SciPy-user] scipy workflow
In-Reply-To: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
Message-ID: <48E14401.7030600@ru.nl>



Jo?o Quinta da Fonseca wrote:
> I have given up Matlab and I it feels great to use scipy for my  
> scientific computing needs. However, I find that I use scipy just like  
> I used matlab, with ipython as my terminal. Although this works OK,  
> some things are a little frustrating: updated modules need to be  
> reimported after modification, running scripts requires the run  
> command etc., but this is to be expected because scipy is not matlbab.
> Now I don't want the matlab way necessarily and I am happy to learn a  
> new way to do things. I guess what I would like to do is do things the  
> python way, which I think means writing my code as classes with the  
> _main_ bit at the end etc., but I am not sure whether that is the best  
> way and if so, what is the best way to learn it.
> Does any one feel like sharing what their workflow with scipy is? Any  
> tips for me?
>
>   
For now I use Signal WorkBench as a MatLab replacement,
it misses the interactive single line statement and MatLab's workspace.
Changed modules are always automatically reloaded.
   
http://oase.uci.kun.nl/~mientki/data_www/pylab_works/pw_signal_workbench.html

And here is concrete example of filter design, compared to how we did it 
in MatLab
  http://oase.uci.kun.nl/~mientki/data_www/pylab_works/pw_filter_design.html

As part of the project I'm now improving the editor / debugger,
which will both solve above mentioned omissions.
As soon as this works well, it's almost also implemented in Signal 
WorkBench.
  http://oase.uci.kun.nl/~mientki/data_www/pylab_works/pw_debug.html

cheers,
Stef
> Thanks,
>
> Joao
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>   


From lo.maximo73 at gmail.com  Tue Sep 30 00:43:10 2008
From: lo.maximo73 at gmail.com (luis cota)
Date: Mon, 29 Sep 2008 23:43:10 -0500
Subject: [SciPy-user] Problems Building / Installing SciPy on OSX Leopard
Message-ID: <2e598d7f0809292143x508cec5dq806bd3136dfd8987@mail.gmail.com>

I first tried installing using the OSX SuperPack.  The install seemed to
work, though when importing scipy.integrate, Python crashed because of an
error with the referenced modules.  This probably has something to do with
my OSX setup, though I am not sure where to look.  I have installed the
latest gFortran and the latest FFTW-3 library as well.  I have appended my
build log at the end of this email.

Any help greatly appreciated ... thanks!


*When compiling the SVN Trunk version I get the following output:
*

python setup.py build
Warning: No configuration returned, assuming unavailable.
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
  libraries mkl,vml,guide not found in /opt/local/lib
  NOT AVAILABLE

fftw3_info:
  libraries fftw3 not found in
/Library/Frameworks/Python.framework/Versions/2.5/lib
  libraries fftw3 not found in /usr/local/lib
  libraries fftw3 not found in /usr/lib
  FOUND:
    libraries = ['fftw3']
    library_dirs = ['/opt/local/lib']
    define_macros = [('SCIPY_FFTW3_H', None)]
    include_dirs = ['/opt/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 = ['-msse3',
'-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 = ['-msse3']

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
  libraries umfpack not found in /opt/local/lib
/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/numpy-1.3.0.dev5861-py2.5-macosx-10.3-i386.egg/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

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 "odrpack" sources
building library "minpack" sources
building library "rootfind" sources
building library "superlu_src" sources
building library "arpack" 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.cluster._distance_wrap" sources
building extension "scipy.cluster._hierarchy_wrap" sources
building extension "scipy.fftpack._fftpack" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.fftpack.convolve" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' 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-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.interpolate._fitpack" sources
building extension "scipy.interpolate.dfitpack" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
  adding
'build/src.macosx-10.3-i386-2.5/scipy/interpolate/src/dfitpack-f2pywrappers.f'
to sources.
building extension "scipy.interpolate._interpolate" sources
building extension "scipy.io.numpyio" sources
building extension "scipy.lib.blas.fblas" sources
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
  adding
'build/src.macosx-10.3-i386-2.5/build/src.macosx-10.3-i386-2.5/scipy/lib/blas/fblas-f2pywrappers.f'
to sources.
building extension "scipy.lib.blas.cblas" sources
  adding 'build/src.macosx-10.3-i386-2.5/scipy/lib/blas/cblas.pyf' to
sources.
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.lib.lapack.flapack" sources
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.lib.lapack.clapack" sources
  adding 'build/src.macosx-10.3-i386-2.5/scipy/lib/lapack/clapack.pyf' to
sources.
f2py options: ['skip:', ':']
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.lib.lapack.calc_lwork" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.lib.lapack.atlas_version" sources
building extension "scipy.linalg.fblas" sources
  adding 'build/src.macosx-10.3-i386-2.5/scipy/linalg/fblas.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
  adding
'build/src.macosx-10.3-i386-2.5/build/src.macosx-10.3-i386-2.5/scipy/linalg/fblas-f2pywrappers.f'
to sources.
building extension "scipy.linalg.cblas" sources
  adding 'build/src.macosx-10.3-i386-2.5/scipy/linalg/cblas.pyf' to sources.
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.linalg.flapack" sources
  adding 'build/src.macosx-10.3-i386-2.5/scipy/linalg/flapack.pyf' to
sources.
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
  adding
'build/src.macosx-10.3-i386-2.5/build/src.macosx-10.3-i386-2.5/scipy/linalg/flapack-f2pywrappers.f'
to sources.
building extension "scipy.linalg.clapack" sources
  adding 'build/src.macosx-10.3-i386-2.5/scipy/linalg/clapack.pyf' to
sources.
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.linalg._flinalg" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.linalg.calc_lwork" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.linalg.atlas_version" 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-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.optimize.moduleTNC" sources
building extension "scipy.optimize._cobyla" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.optimize.minpack2" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.optimize._slsqp" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.optimize._nnls" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.signal.sigtools" sources
building extension "scipy.signal.spline" sources
building extension "scipy.sparse.linalg.isolve._iterative" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.sparse.linalg.dsolve._zsuperlu" sources
building extension "scipy.sparse.linalg.dsolve._dsuperlu" sources
building extension "scipy.sparse.linalg.dsolve._csuperlu" sources
building extension "scipy.sparse.linalg.dsolve._ssuperlu" sources
building extension "scipy.sparse.linalg.dsolve.umfpack.__umfpack" sources
building extension "scipy.sparse.linalg.eigen.arpack._arpack" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
  adding
'build/src.macosx-10.3-i386-2.5/build/src.macosx-10.3-i386-2.5/scipy/sparse/linalg/eigen/arpack/_arpack-f2pywrappers.f'
to sources.
building extension "scipy.sparse.sparsetools._csr" sources
building extension "scipy.sparse.sparsetools._csc" sources
building extension "scipy.sparse.sparsetools._coo" sources
building extension "scipy.sparse.sparsetools._bsr" sources
building extension "scipy.sparse.sparsetools._dia" sources
building extension "scipy.special._cephes" sources
building extension "scipy.special.specfun" sources
f2py options: ['--no-wrap-functions']
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.stats.statlib" sources
f2py options: ['--no-wrap-functions']
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.stats.futil" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
building extension "scipy.stats.mvn" sources
f2py options: []
  adding 'build/src.macosx-10.3-i386-2.5/fortranobject.c' to sources.
  adding 'build/src.macosx-10.3-i386-2.5' to include_dirs.
  adding 'build/src.macosx-10.3-i386-2.5/scipy/stats/mvn-f2pywrappers.f' to
sources.
building extension "scipy.ndimage._nd_image" 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 build/src.macosx-10.3-i386-2.5/scipy/__config__.py ->
build/lib.macosx-10.3-i386-2.5/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
building 'dfftpack' library
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -fPIC -O3 -funroll-loops
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second-underscore
-fPIC -O3 -funroll-loops
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -Wall -fno-second-underscore -fPIC -O3 -funroll-loops
creating build/temp.macosx-10.3-i386-2.5
creating build/temp.macosx-10.3-i386-2.5/scipy
creating build/temp.macosx-10.3-i386-2.5/scipy/fftpack
creating build/temp.macosx-10.3-i386-2.5/scipy/fftpack/dfftpack
compile options: '-c'
gfortran:f77: scipy/fftpack/dfftpack/dcosqb.f
gfortran:f77: scipy/fftpack/dfftpack/dcosqf.f
gfortran:f77: scipy/fftpack/dfftpack/dcosqi.f
gfortran:f77: scipy/fftpack/dfftpack/dcost.f
gfortran:f77: scipy/fftpack/dfftpack/dcosti.f
gfortran:f77: scipy/fftpack/dfftpack/dfftb.f
gfortran:f77: scipy/fftpack/dfftpack/dfftb1.f
gfortran:f77: scipy/fftpack/dfftpack/dfftf.f
gfortran:f77: scipy/fftpack/dfftpack/dfftf1.f
gfortran:f77: scipy/fftpack/dfftpack/dffti.f
gfortran:f77: scipy/fftpack/dfftpack/dffti1.f
scipy/fftpack/dfftpack/dffti1.f: In function 'dffti1':
scipy/fftpack/dfftpack/dffti1.f:11: warning: 'ntry' may be used
uninitialized in this function
gfortran:f77: scipy/fftpack/dfftpack/dsinqb.f
gfortran:f77: scipy/fftpack/dfftpack/dsinqf.f
gfortran:f77: scipy/fftpack/dfftpack/dsinqi.f
gfortran:f77: scipy/fftpack/dfftpack/dsint.f
gfortran:f77: scipy/fftpack/dfftpack/dsint1.f
gfortran:f77: scipy/fftpack/dfftpack/dsinti.f
gfortran:f77: scipy/fftpack/dfftpack/zfftb.f
gfortran:f77: scipy/fftpack/dfftpack/zfftb1.f
gfortran:f77: scipy/fftpack/dfftpack/zfftf.f
gfortran:f77: scipy/fftpack/dfftpack/zfftf1.f
gfortran:f77: scipy/fftpack/dfftpack/zffti.f
gfortran:f77: scipy/fftpack/dfftpack/zffti1.f
scipy/fftpack/dfftpack/zffti1.f: In function 'zffti1':
scipy/fftpack/dfftpack/zffti1.f:11: warning: 'ntry' may be used
uninitialized in this function
ar: adding 23 object files to build/temp.macosx-10.3-i386-2.5/libdfftpack.a
ranlib:@ build/temp.macosx-10.3-i386-2.5/libdfftpack.a
building 'linpack_lite' library
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -fPIC -O3 -funroll-loops
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second-underscore
-fPIC -O3 -funroll-loops
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -Wall -fno-second-underscore -fPIC -O3 -funroll-loops
creating build/temp.macosx-10.3-i386-2.5/scipy/integrate
creating build/temp.macosx-10.3-i386-2.5/scipy/integrate/linpack_lite
compile options: '-c'
gfortran:f77: scipy/integrate/linpack_lite/dgbfa.f
gfortran:f77: scipy/integrate/linpack_lite/dgbsl.f
gfortran:f77: scipy/integrate/linpack_lite/dgefa.f
gfortran:f77: scipy/integrate/linpack_lite/dgesl.f
gfortran:f77: scipy/integrate/linpack_lite/dgtsl.f
gfortran:f77: scipy/integrate/linpack_lite/zgbfa.f
gfortran:f77: scipy/integrate/linpack_lite/zgbsl.f
gfortran:f77: scipy/integrate/linpack_lite/zgefa.f
gfortran:f77: scipy/integrate/linpack_lite/zgesl.f
ar: adding 9 object files to
build/temp.macosx-10.3-i386-2.5/liblinpack_lite.a
ranlib:@ build/temp.macosx-10.3-i386-2.5/liblinpack_lite.a
building 'mach' library
using additional config_fc from setup script for fortran compiler: {'noopt':
('scipy/integrate/setup.pyc', 1)}
customize Gnu95FCompiler
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -fPIC
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second-underscore
-fPIC
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -Wall -fno-second-underscore -fPIC
creating build/temp.macosx-10.3-i386-2.5/scipy/integrate/mach
compile options: '-c'
gfortran:f77: scipy/integrate/mach/d1mach.f
gfortran:f77: scipy/integrate/mach/i1mach.f
gfortran:f77: scipy/integrate/mach/r1mach.f
gfortran:f77: scipy/integrate/mach/xerror.f
scipy/integrate/mach/xerror.f:1.40:

      SUBROUTINE XERROR(MESS,NMESS,L1,L2)
                                       1
Warning: Unused dummy argument 'l2' at (1)
scipy/integrate/mach/xerror.f:1.37:

      SUBROUTINE XERROR(MESS,NMESS,L1,L2)
                                    1
Warning: Unused dummy argument 'l1' at (1)
ar: adding 4 object files to build/temp.macosx-10.3-i386-2.5/libmach.a
ranlib:@ build/temp.macosx-10.3-i386-2.5/libmach.a
building 'quadpack' library
compiling Fortran sources
Fortran f77 compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -fPIC -O3 -funroll-loops
Fortran f90 compiler: /usr/local/bin/gfortran -Wall -fno-second-underscore
-fPIC -O3 -funroll-loops
Fortran fix compiler: /usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -Wall -fno-second-underscore -fPIC -O3 -funroll-loops
creating build/temp.macosx-10.3-i386-2.5/scipy/integrate/quadpack
compile options: '-c'
gfortran:f77: scipy/integrate/quadpack/dqag.f
gfortran:f77: scipy/integrate/quadpack/dqage.f
gfortran:f77: scipy/integrate/quadpack/dqagi.f
gfortran:f77: scipy/integrate/quadpack/dqagie.f
scipy/integrate/quadpack/dqagie.f: In function 'dqagie':
scipy/integrate/quadpack/dqagie.f:154: warning: 'small' may be used
uninitialized in this function
scipy/integrate/quadpack/dqagie.f:153: warning: 'ertest' may be used
uninitialized in this function
scipy/integrate/quadpack/dqagie.f:152: warning: 'erlarg' may be used
uninitialized in this function
scipy/integrate/quadpack/dqagie.f:151: warning: 'correc' may be used
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqagp.f
gfortran:f77: scipy/integrate/quadpack/dqagpe.f
scipy/integrate/quadpack/dqagpe.f: In function 'dqagpe':
scipy/integrate/quadpack/dqagpe.f:196: warning: 'k' may be used
uninitialized in this function
scipy/integrate/quadpack/dqagpe.f:191: warning: 'correc' may be used
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqags.f
gfortran:f77: scipy/integrate/quadpack/dqagse.f
scipy/integrate/quadpack/dqagse.f: In function 'dqagse':
scipy/integrate/quadpack/dqagse.f:153: warning: 'small' may be used
uninitialized in this function
scipy/integrate/quadpack/dqagse.f:152: warning: 'ertest' may be used
uninitialized in this function
scipy/integrate/quadpack/dqagse.f:151: warning: 'erlarg' may be used
uninitialized in this function
scipy/integrate/quadpack/dqagse.f:150: warning: 'correc' may be used
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqawc.f
gfortran:f77: scipy/integrate/quadpack/dqawce.f
gfortran:f77: scipy/integrate/quadpack/dqawf.f
gfortran:f77: scipy/integrate/quadpack/dqawfe.f
scipy/integrate/quadpack/dqawfe.f: In function 'dqawfe':
scipy/integrate/quadpack/dqawfe.f:203: warning: 'll' may be used
uninitialized in this function
scipy/integrate/quadpack/dqawfe.f:200: warning: 'drl' may be used
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqawo.f
gfortran:f77: scipy/integrate/quadpack/dqawoe.f
scipy/integrate/quadpack/dqawoe.f: In function 'dqawoe':
scipy/integrate/quadpack/dqawoe.f:208: warning: 'ertest' may be used
uninitialized in this function
scipy/integrate/quadpack/dqawoe.f:207: warning: 'erlarg' may be used
uninitialized in this function
scipy/integrate/quadpack/dqawoe.f:206: warning: 'correc' may be used
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqaws.f
gfortran:f77: scipy/integrate/quadpack/dqawse.f
gfortran:f77: scipy/integrate/quadpack/dqc25c.f
gfortran:f77: scipy/integrate/quadpack/dqc25f.f
scipy/integrate/quadpack/dqc25f.f: In function 'dqc25f':
scipy/integrate/quadpack/dqc25f.f:103: warning: 'm' may be used
uninitialized in this function
gfortran:f77: scipy/integrate/quadpack/dqc25s.f
gfortran:f77: scipy/integrate/quadpack/dqcheb.f
gfortran:f77: scipy/integrate/quadpack/dqelg.f
scipy/integrate/quadpack/dqelg.f: In function 'dqelg':
scipy/integrate/quadpack/dqelg.f:1: internal compiler error: vector
VEC(tree,base) index domain error, in build_classic_dist_vector_1 at
tree-data-ref.c:2725
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
scipy/integrate/quadpack/dqelg.f: In function 'dqelg':
scipy/integrate/quadpack/dqelg.f:1: internal compiler error: vector
VEC(tree,base) index domain error, in build_classic_dist_vector_1 at
tree-data-ref.c:2725
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
error: Command "/usr/local/bin/gfortran -Wall -ffixed-form
-fno-second-underscore -fPIC -O3 -funroll-loops -c -c
scipy/integrate/quadpack/dqelg.f -o
build/temp.macosx-10.3-i386-2.5/scipy/integrate/quadpack/dqelg.o" failed
with exit status 1
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080929/e25ad967/attachment.html>

From david at ar.media.kyoto-u.ac.jp  Tue Sep 30 00:35:56 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 30 Sep 2008 13:35:56 +0900
Subject: [SciPy-user] Problems Building / Installing SciPy on OSX Leopard
In-Reply-To: <2e598d7f0809292143x508cec5dq806bd3136dfd8987@mail.gmail.com>
References: <2e598d7f0809292143x508cec5dq806bd3136dfd8987@mail.gmail.com>
Message-ID: <48E1ACAC.9070907@ar.media.kyoto-u.ac.jp>

luis cota wrote:
> I first tried installing using the OSX SuperPack.  The install seemed
> to work, though when importing scipy.integrate, Python crashed because
> of an error with the referenced modules.  This probably has something
> to do with my OSX setup, though I am not sure where to look.  I have
> installed the latest gFortran and the latest FFTW-3 library as well. 
> I have appended my build log at the end of this email.
>
> Any help greatly appreciated ... thanks!

Hi Luis,

    I am sorry for the trouble. For the building part, which gfortran
are you using ? I suspect you are using a very recent gfortran, which is
not stable (you see a compiler crash, which is not a good sign :) ).

Please use this one:

http://r.research.att.com/tools/

In particular, do not use this one for the time being:

http://hpc.sourceforge.net/

cheers,

David


From lo.maximo73 at gmail.com  Tue Sep 30 01:50:02 2008
From: lo.maximo73 at gmail.com (luis cota)
Date: Tue, 30 Sep 2008 00:50:02 -0500
Subject: [SciPy-user] Problems Building / Installing SciPy on OSX Leopard
In-Reply-To: <48E1ACAC.9070907@ar.media.kyoto-u.ac.jp>
References: <2e598d7f0809292143x508cec5dq806bd3136dfd8987@mail.gmail.com>
	<48E1ACAC.9070907@ar.media.kyoto-u.ac.jp>
Message-ID: <2e598d7f0809292250s5ad33320o9d0a3ddd65e2d1a@mail.gmail.com>

Thanks! That was the problem - after installing the ATT version of gFortran
the build worked fine.

- Luis

On Mon, Sep 29, 2008 at 11:35 PM, David Cournapeau <
david at ar.media.kyoto-u.ac.jp> wrote:

> luis cota wrote:
> > I first tried installing using the OSX SuperPack.  The install seemed
> > to work, though when importing scipy.integrate, Python crashed because
> > of an error with the referenced modules.  This probably has something
> > to do with my OSX setup, though I am not sure where to look.  I have
> > installed the latest gFortran and the latest FFTW-3 library as well.
> > I have appended my build log at the end of this email.
> >
> > Any help greatly appreciated ... thanks!
>
> Hi Luis,
>
>    I am sorry for the trouble. For the building part, which gfortran
> are you using ? I suspect you are using a very recent gfortran, which is
> not stable (you see a compiler crash, which is not a good sign :) ).
>
> Please use this one:
>
> http://r.research.att.com/tools/
>
> In particular, do not use this one for the time being:
>
> http://hpc.sourceforge.net/
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080930/4d1a2712/attachment.html>

From cmac at mit.edu  Tue Sep 30 10:09:52 2008
From: cmac at mit.edu (Christopher MacMinn)
Date: Tue, 30 Sep 2008 10:09:52 -0400
Subject: [SciPy-user] sparse matrices -- slicing and fancy indexing for
	assignment
Message-ID: <4FBBE3B4-041E-4593-8ADA-1318BF43A853@mit.edu>

Hey folks -

I installed SciPy 0.7.0.dev4753 from svn yesterday to get access to  
some of the upcoming improvements to scipy.sparse.

In particular, I was excited about the prospect of slicing and fancy  
indexing for CSC and CSR matrices.  Unfortunately, I find now that  
slicing and fancy indexing are only supported for viewing and copying,  
not for assignment.  E.g.:

# Create a sparse CSC matrix A

 >>> from scipy import sparse
 >>> A = sparse.eye(10,10,format="csc")

# Have a look at the first row of A

 >>> print A[0,:].todense()
[[ 1.  0.  0.  0.  0.  0.  0.  0.  0.  0.]]

# Now try to set the first row equal to 5

 >>> A[0,:] = 5.
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
   File "/Library/Python/2.5/site-packages/scipy/sparse/ 
compressed.py", line 601, in __setitem__
     raise NotImplementedError("Fancy indexing in assignment not "
NotImplementedError: Fancy indexing in assignment not supported for  
csr matrices.

# Fail.

Is slicing / fancy indexing for assignment planned for the final 0.7.0  
release, or is this (above) the best we're going to get?

Best, Chris MacMinn


From wnbell at gmail.com  Tue Sep 30 12:01:06 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Tue, 30 Sep 2008 12:01:06 -0400
Subject: [SciPy-user] sparse matrices -- slicing and fancy indexing for
	assignment
In-Reply-To: <4FBBE3B4-041E-4593-8ADA-1318BF43A853@mit.edu>
References: <4FBBE3B4-041E-4593-8ADA-1318BF43A853@mit.edu>
Message-ID: <d05265cb0809300901i2a20facam421af339cb8238b7@mail.gmail.com>

On Tue, Sep 30, 2008 at 10:09 AM, Christopher MacMinn <cmac at mit.edu> wrote:
>
> Is slicing / fancy indexing for assignment planned for the final 0.7.0
> release, or is this (above) the best we're going to get?
>

This is probably all you'll see in 0.7.0

Note that assigning to a CSR/CSC matrix is in general a bad idea.  Any
change to the sparsity structure of these formats requires O(nnz)
operations, which basically means reconstructing the matrix from
scratch.

OTOH MATLAB lets you do it, so we will probably support it someday
too.  If you submit a patch soon we might be able to integrate it by
0.7.0.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From flyingdeckchair at googlemail.com  Tue Sep 30 12:36:00 2008
From: flyingdeckchair at googlemail.com (peter websdell)
Date: Tue, 30 Sep 2008 17:36:00 +0100
Subject: [SciPy-user] unusual fitting problem
Message-ID: <b4246c960809300936q325f2e75ubed1c90582a45025@mail.gmail.com>

Howdy gang,
I have an unusual fitting problem that has me totally stumped.

I need to perform a basic interpolation between a few data points, but i
need the result to look linear on a loglog scale. Making sense?

OK, i'll try again. Imagine you plotted the data points on a loglog scale,
then printed it out and joined the dots with a ruler and pencil. That is
what I need to achieve.

I'm trying to automate an old assessment method at work which requires
manually drawing a chart to extrapolate new values between the known data
points. It's tedious and there are literally hundreds need doing. :-o

Any help will be graciously received. Please let me know if my explanation
is nonsense.

Cheers,
Pete.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20080930/8118abe0/attachment.html>

From pgmdevlist at gmail.com  Tue Sep 30 12:40:54 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 30 Sep 2008 12:40:54 -0400
Subject: [SciPy-user] unusual fitting problem
In-Reply-To: <b4246c960809300936q325f2e75ubed1c90582a45025@mail.gmail.com>
References: <b4246c960809300936q325f2e75ubed1c90582a45025@mail.gmail.com>
Message-ID: <200809301240.54583.pgmdevlist@gmail.com>

On Tuesday 30 September 2008 12:36:00 peter websdell wrote:
> Howdy gang,
> I have an unusual fitting problem that has me totally stumped.

Have you thought about modifying your data ?
you want log(y) = a + b*log(x) ?
Use Y=log(y) and X=log(X), fit a straight-line the standard way and you should 
get your parameters a & b.
That should be easier than trying to fit y=exp(a+b*log(x)) = exp(a) * x**b



From aisaac at american.edu  Tue Sep 30 13:09:32 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Tue, 30 Sep 2008 13:09:32 -0400
Subject: [SciPy-user] NumpyTest warning after upgrade
In-Reply-To: <1d36917a0809281236r5042b66fxddee186e7ec405d2@mail.gmail.com>
References: <88fe22a0809281205p11155ff8l2204ffa5d63011ff@mail.gmail.com>
	<1d36917a0809281236r5042b66fxddee186e7ec405d2@mail.gmail.com>
Message-ID: <48E25D4C.9000000@american.edu>

On 9/28/2008 3:36 PM Alan McIntyre apparently wrote:
> if you just want to patch up your local install, you 
> should be able to change code that looks like this in 
> scipy/misc/__init__.py (and probably several other 
> places):

> from numpy.testing import NumpyTest 
> test = NumpyTest().test 

> to this:

> from numpy.testing import Tester 
> test = Tester().test 


About a dozen other places.
But this has been working for me.

Cheers,
Alan Isaac