Ian Henriksen | 7 Sep 03:09 2014
Picon

Expose Lapack for Cython

Hi all, I have put together pxd files that allow direct access from Cython to the the blas and lapack functions wrapped in scipy.linalg.blas and scipy.linalg.lapack. The idea is to make these low level functions available in Cython without the corresponding Python level function call that is necessary when using the wrappers already in SciPy. They are currently in my github repository https://github.com/insertinterestingnamehere/cylinalg. The approach is based on the gist https://gist.github.com/pv/5437087. I think this would be a good addition to SciPy since it is really only another interface to the wrappers that are already there.

Where would be the best place to put this? Also, what sorts of tests should be included? Any other input is welcome too.

Thanks!

-Ian Henriksen
_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
Clark Fitzgerald | 6 Sep 18:54 2014
Picon

return named tuples

As a user, it would be convenient to access results by name. The relevant issue on Github is #3665. For example:

summary = scipy.stats.describe(data)

The current way to access the mean is to write `summary[2]`. It would be more clear to write `summary.mean`. I'd like to create a PR to implement this using namedtuples. Here is the idea:

from collections import namedtuple
from scipy.stats import describe, norm


data = norm().rvs(100)
unnamed = describe(data)

output = namedtuple('describe', ('size', 'range', 'mean', 'variance',
                                 'skewness', 'kurtosis'))

named = output(*unnamed)

This doesn't break backwards compatibility since:

In [8]: unnamed[2] == named[2] == named.mean
Out[8]: True

This isn't limited to stats.describe; it would be beneficial anywhere that the return object is a tuple that's possibly unclear. 

The feedback on the Git issue has been positive. More feedback welcome.


Best,
Clark Fitzgerald
Statistics phd student, UC Davis
_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
Jaime Fernández del Río | 5 Sep 22:17 2014
Picon

bug in min(max)imum_filter1d

A while back I made some changes to ndimage.filters.min(max)imum_filter1d that sped it up quite considerably, see here:


A little after that, I started a discussion on whether a different algorithm, about 30% slower for random inputs, but up to 10x faster for worse case inputs, would be a better option, and submitted a new PR, see here:


I didn't follow up on that discussion, and so #3527 had been lingering abandoned ever since.

Recently, Julian Taylor discovered that #3517 breaks the scikit-image tests, because it does not handle properly the case with a filter of size 1, see here:


Rather than fixing the existing code merged in #3517, I have reworked the code in #3527 a little, because it had the same trouble with filters of size 1, and updated it. Merging the latest version of #3527 will fix the skimage bug, but I would like to bring up the discussion on whether this is the right thing to do or not. I certainly prefer #3527 to #3517, but if there is no consensus I will close #3527 and fix #3517 instead.

Thanks,

Jaime

--
(\__/)
( O.o)
( > <) Este es Conejo. Copia a Conejo en tu firma y ayúdale en sus planes de dominación mundial.
_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
Ralf Gommers | 5 Sep 21:07 2014
Picon

0.14.1 and 0.15.0 release schedule

Hi all,

It's about time for a couple of releases.

0.14.1 fixes some regressions as well as a lot of test failures against numpy 1.9.0 (due to __numpy_ufunc__) removal. All known regressions are fixed now, so I plan to tag it this weekend. If there's anything really urgent that still needs to go in, please speak up.

For 0.15.0 there are still 17 issues/PRs, but most of those we should be able to merge/fix relatively quickly. I propose the following release schedule:

- beta 1: 16 Sep
- release candidate 1: 4 Oct
- release candidate 2: 18 Oct (if needed)
- final release: 25 Oct

Does that work for everyone? Are there issues/PRs that aren't under the 0.15.0 milestone but have to go in?

Cheers,
Ralf


_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
David J Pine | 5 Sep 16:25 2014
Picon

Introduction to Python for Science

NumPy, SciPy, MatPlotLib Users & Science teachers:

I have written an introduction to scientific python that you may find useful.  You can download it from GitHub and use it freely:


I wrote this manual/book for undergraduates taking science and engineering courses that use programming to solve science and engineering problems.  It is not for experts.  I am sharing it with the hope that others may find it useful.  It includes an introduction to very basic programming, numpy,matplotlib, & scipy, as well as instructions on how to download and install Python and these three libraries.  It also includes an introduction to IPython notebooks.

Corrections and suggestions for improvements are welcome.

David Pine

_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
Joseph Martinot-Lagarde | 4 Sep 01:58 2014

Multiple comment tokens for loadtxt

loadtxt currently has a keyword to change the comment token. The PR 
#4612 [1] enables to define multiple comment token for a file. It is 
motivated by #2633 [2]

What is your position on this one ?

Joseph

     [1] https://github.com/numpy/numpy/pull/4612
     [2] https://github.com/numpy/numpy/issues/2633
Joseph Martinot-Lagarde | 4 Sep 01:51 2014

'norm' keyword for FFT functions

I have an old PR [1] to fix #2142 [2]. The idea is to have a new keyword 
for all fft functions to define the normalisation of the fft:
- if 'norm' is None (the default), the normalisation is the current one: 
fft() is not normalized ans ifft is normalized by 1/n.
- if norm is "ortho", the direct and inverse transforms are both 
normalized by 1/sqrt(n). The results are then unitary.

The keyword name and value is consistent with scipy.fftpack.dct.

Do you feel that it should be merged ?

Joseph

     [1] https://github.com/numpy/numpy/pull/3883
     [2] https://github.com/numpy/numpy/issues/2142
Ralf Gommers | 1 Sep 12:36 2014
Picon

EuroSciPy sprint report

Hi all,

We had a great sprint at EuroSciPy yesterday, with about 20 people working on Scipy and about 25 more on scikit-learn, scikit-image and Vispy. The result at the end of the day for Scipy was:

  - 24 pull requests submitted
  - 20 pull requests merged
  - 17 issues closed
  - 2 new issues created

A couple of highlights:

  - Fred Ludlow implemented bitwise operations on sparse matrices (PR 3941)
  - Julian Taylor implemented releasing the GIL from filter and interpolation functions in ndimage (PR 3943)
  - Aaren O'Leary implemented API changes needed before integrating discrete wavelets into scipy (issue 3931).
  - Chris Kerr removed linpack_lite from scipy.integrate (PR 3916). (done right before the sprint actually, but removing 15000 lines of Fortran code is worth a mention anyway!).

It was great to see a lot of new contributors as well as meet people who have been contributing for a while face-to-face. Thanks a lot to everyone who contributed, and especially to Pietro for the excellent organization!

Cheers,
Ralf


_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
Moritz Beber | 29 Aug 12:13 2014
Picon

nested setup.py scripts

Dear all,

I want to generate a package with a submodule structure similar to what numpy and scipy use. (Or do you recommend not doing that?)  I have read the following pieces of documentation but I'm still unclear about how the main setup.py script discovers the nested scripts and gets the configuration values from those. Is this documented somewhere or can anyone point me to how this is done?

Thank you in advance,
Moritz

P.S.: What I've read:
https://github.com/numpy/numpy/blob/master/doc/DISTUTILS.rst.txt
http://docs.scipy.org/doc/scipy-dev/reference/hacking.html
http://docs.scipy.org/doc/scipy/reference/api.html
_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
Sri Krishna | 26 Aug 14:36 2014
Picon

To use C code or Cython code?

Hi,

I'm new to the Scipy-Dev mailing list, looking to contribute wherever I can. I was looking through the open issues and saw this issue, regarding a speed-up for the convolve2d function.

My confusion arises from the SciPy coding guidelines which states that using Cython is much preferable to using plain C/C++/Fortran.

Would it be desirable then to change the C code of signal/firfilter.c to a Cythonized code?

Thanks,
Krishna

_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev
Jeff Grasty | 26 Aug 00:55 2014
Picon

Nyquist Filters

Hi,

One of the features that I have found missing in SciPy are functions to design nyquist and root-nyquist
filters, such as raised cosine and root-raised cosine filters.  I have written several functions for this
purpose and was curious if anyone thought was a greater need for this. 

Thanks,
Jeff
_______________________________________________
SciPy-Dev mailing list
SciPy-Dev <at> scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-dev

Gmane