Mailman 3 June 2014 - NumPy-Discussion

SciPy 2014 BoF NumPy Participation
by Kyle Mandli June 29, 2014

June 29, 2014

Hello everyone, As one of the co-chairs in charge of organizing the birds-of-a-feather sesssions at the SciPy conference this year, I wanted to solicit through the NumPy list to see if we could get enough interest to hold a NumPy centered BoF this year. The BoF format would be up to those who would lead the discussion, a couple of ideas used in the past include picking out a few of the lead devs to be on a panel and have a Q&A type of session or an open Q&A with perhaps audience guided list of topics. I can help facilitate organization of something but we would really like to get something organized this year (last year NumPy was the only major project that was not really represented in the BoF sessions). Thanks! Kyle Manldi (and via proxy Matt McCormick)

10 17

NumPy en français
by pi8L4vion June 23, 2014

June 23, 2014

Bonsoir à tous, Est-ce quil y a des francophones qui pourraient échanger avec moi ? Merci. Pi8L4vion

2 1

Find n closest values
by Nicolas P. Rougier June 23, 2014

June 23, 2014

Hi, I have an array L with regular spaced values between 0 and width. I have a (sorted) array I with irregular spaced values between 0 and width. I would like to find the closest value in I for any value in L. Currently, I'm using the following script but I wonder if I missed an obvious (and faster) solution: import numpy as np def find_closest(A, target): idx = A.searchsorted(target) idx = np.clip(idx, 1, len(A) - 1) left = A[idx - 1] right = A[idx] idx -= target - left < right - target return idx n, width = 256, 100.0 # 10 random sorted values in [0,width] I = np.sort(np.random.randint(0,width,10)) # n regular spaced values in [0,width] L = np.linspace(0, width, n) print I[find_closest(I,L)] Nicolas

4 11

100 numpy exercises, now in Julia
by Nicolas P. Rougier June 22, 2014

June 22, 2014

Hi all, Michiaki Ariga has started conversion of 100 numpy exercises in Julia. I know this might be a bit off-topic but I thought it was interesting enough. Github repository is at https://github.com/chezou/julia-100-exercises Nicolas

1 0

Conditionally compiling Fortran 90 extensions using numpy.distutils
by Jonathan Helmus June 20, 2014

June 20, 2014

I have been working on creating a setup.py file for a sub-module with the aim to conditionally compile a Fortran 90 extension using F2PY depending on the availability of an appropriate compiler on the system. Basically if gfortran or another Fortran 90 compiler is available on the system, the extension should be built. If no compiler is available, the extension should not be built and the build should continue without error. After a bit of work I was able to get a working script which uses numpy.distutils: from os.path import join def configuration(parent_package='', top_path=None): global config from numpy.distutils.misc_util import Configuration config = Configuration('retrieve', parent_package, top_path) config.add_data_dir('tests') # Conditionally add Steiner echo classifier extension. config.add_extension('echo_steiner',sources=[steiner_echo_gen_source]) return config def steiner_echo_gen_source(ext, build_dir): try: config.have_f90c() return [join(config.local_path, 'echo_steiner.pyf'), join(config.local_path, 'src', 'echo_steiner.f90')] except: return None if __name__ == '__main__': from numpy.distutils.core import setup setup(**configuration(top_path='').todict()) Is there a better way of accomplishing this conditional compiling, perhaps without using a global variable or a try/except block? Additionally is the expected behaviour of the have_90c function to raise an exception when a Fortran 90 compiler is not available or is this a bug? From the documentation [1] it was unclear what the results should be. I should mention that the above code snippet was aided greatly by information in the NumPy Packinging documentation [2], NumPy Distutils - Users Guide [3], and code from the f2py utils unit tests [4]. Thanks, - Jonathan Helmus nmrglue.com/jhelmus [1] http://docs.scipy.org/doc/numpy/reference/distutils.html#numpy.distutils.mi… [2] http://docs.scipy.org/doc/numpy/reference/distutils.html [3] http://wiki.scipy.org/Wiki/Documentation/numpy_distutils [4] https://github.com/numpy/numpy/blob/master/numpy/f2py/tests/util.py

1 0

Automated test build machinery on travis - move to numpy or scipy org?
by Matthew Brett June 19, 2014

June 19, 2014

Hi, As some of you might have seen, I automated the build of numpy / scipy against ATLAS using recent gcc here: https://github.com/matthew-brett/numpy-atlas-binaries with testing and uploading to Rackspace here: https://travis-ci.org/matthew-brett/numpy-atlas-binaries (thanks to Matt Terry for the original code for travis-ci build / testing [1] and Olivier Grisel for setting me up for Rackspace upload). Maybe it would be a good idea to move numpy-atlas-build to numpy or scipy organization and maintain it as part of the release process? Cheers, Matthew [1] https://github.com/matplotlib/mpl_mac_testing

1 0

Switch to using ATLAS for OSX binary wheels
by Matthew Brett June 19, 2014

June 19, 2014

Hi, Summary : I'm planning to upload OSX wheels for numpy and scipy using the ATLAS blas / lapack library instead of the default OSX Accelerate framework. We've run into some trouble with a segfault in recent OSX Accelerate: https://github.com/numpy/numpy/issues/4007 and Accelerate also doesn't play well with multiprocessing. https://github.com/numpy/numpy/issues/4776 Because there's nothing I love more than half-day compilation runs on my laptop, I've built ATLAS binaries with gcc 4.8, and linked numpy and scipy to them to make OSX wheels. These pass all tests in i386 and x86_64 mode, including numpy, scipy, matplotlib, pandas: https://travis-ci.org/matthew-brett/scipy-stack-osx-testing/builds/27442987 The build process needs some automating, but it's recorded here: https://github.com/matthew-brett/numpy-atlas-binaries It's possible to get travis-ci to build these guys from a bare machine and then upload them somewhere, but I haven't tried to do that. Meanwhile Sturla kindly worked up a patch to numpy to work round the Accelerate segfault [1]. I haven't tested that, but given I'd already built the wheels, I prefer the ATLAS builds because they work with multiprocessing. I propose uploading these wheels as the default for numpy and scipy. Does anyone have any objection or comments before I go ahead and do that? Cheers, Matthew [1] http://mail.scipy.org/pipermail/numpy-discussion/2014-June/070333.html

8 19

numpy.sort() failing mysteriously
by Manolo Martínez June 19, 2014

June 19, 2014

Dear all, I have a couple of structured arrays that, as far as I can see, only differ in irrelevant ways; yet numpy.sort chokes on one, not the other: This structured array *cannot* be sorted with np.sort(array, order='weight')... array([([0, 0, 0], 0.0), ([0, 0, 1], 0.0), ([0, 0, 2], 0.0), ([0, 1, 0], 0.0), ([0, 1, 1], 0.0), ([0, 1, 2], 0.0), ([0, 2, 0], 0.0), ([0, 2, 1], 0.0), ([0, 2, 2], 0.0), ([1, 0, 0], 0.0), ([1, 0, 1], 0.0), ([1, 0, 2], 0.0), ([1, 1, 0], 0.0), ([1, 1, 1], 0.0), ([1, 1, 2], 0.0), ([1, 2, 0], 0.0), ([1, 2, 1], 0.0), ([1, 2, 2], 0.0), ([2, 0, 0], 4.08179211371555e-289), ([2, 0, 1], 0.0), ([2, 0, 2], 0.0), ([2, 1, 0], 1.0), ([2, 1, 1], 6.939504595227983e-225), ([2, 1, 2], 1.0626819127933375e-224), ([2, 2, 0], 1.1209583874093894e-260), ([2, 2, 1], 0.0), ([2, 2, 2], 0.0)], dtype=[('strat', 'O'), ('weight', '<f8')]) ... and this one can array([ ([[0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0.0, 0.0]], -2.40235968796357e-21), ([[0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [1.0, 0.0, 0.0]], -3.3085614943210732e-21), ([[0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [2.0, 0.0, 0.0]], -1.2140117681585496e-23), ..., ([[2.0, 1.0, 0.0], [2.0, 1.0, 0.0], [2.0, 0.5, 0.5]], -1.154494002571172e-21), ([[2.0, 1.0, 0.0], [2.0, 1.0, 0.0], [2.0, 0.75, 0.25]], -1.3867815814699957e-22), ([[2.0, 1.0, 0.0], [2.0, 1.0, 0.0], [2.0, 1.0, 0.0]], 8.722432986511066e-20)], dtype=[('strat', 'O'), ('weight', '<f8')]) The first one leaves the following trace: In [111]: np.sort(aa, order=['weight', 'strat']) --------------------------------------------------------------------------- ValueError Traceback (most recent call last) <ipython-input-111-c3a3525a7694> in <module>() ----> 1 np.sort(aa, order=['weight', 'strat']) /usr/lib/python3.4/site-packages/numpy/core/fromnumeric.py in sort(a, axis, kind, order) 786 else: 787 a = asanyarray(a).copy() --> 788 a.sort(axis, kind, order) 789 return a 790 ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() I should say that I asked this very question earlier today over at #scipy, and jtaylor kindly run the offending array, and could not reproduce the problem. I'm a bit stumped. python --version: 3.4.1 numpy.__version__: 1.8.1 ipython --version: 2.1.0 Thanks for any advice. Cheers, Manolo

1 0

Fail to build rpm using setup.py
by HongQi June 18, 2014

June 18, 2014

Using `python setup.py bdist_rpm`, we can build a rpm for a python package. But when using this command to build rpm for numpy, we get the the following error message: ``` Installed (but unpackaged) file(s) found: /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_api.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_api.pyo /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_arrayprint.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_arrayprint.pyo /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_blasdot.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_blasdot.pyo /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_datetime.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_datetime.pyo /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_defchararray.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_defchararray.pyo /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_dtype.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_dtype.pyo /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_einsum.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_einsum.pyo /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_errstate.pyc /usr/local/lib/python2.7/site-packages/numpy/core/tests/test_errstate.pyo ... ``` and ``` File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_clib.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_clib.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_clib.pyo File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_data.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_data.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_data.pyo File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_headers.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_headers.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/install_headers.pyo File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/scons.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/scons.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/scons.pyo File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/sdist.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/sdist.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/command/sdist.pyo File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/compat.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/compat.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/compat.pyo File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/conv_template.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/conv_template.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/conv_template.pyo File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/core.py File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/core.pyc File listed twice: /usr/local/lib/python2.7/site-packages/numpy/distutils/core.pyo ... ``` we tested building numpy-1.4.1 and numpy-1.7.1 on Redhat 5.7 and building numpy-1.7.1 and numpy-1.8.1 on Fedora 20, none of these builds succeeded, But we can build other python packages(Scrapy) on these system. Any suggestions for this problem? Thanks for your help. Regards, Qiping -- View this message in context: http://numpy-discussion.10968.n7.nabble.com/Fail-to-build-rpm-using-setup-p… Sent from the Numpy-discussion mailing list archive at Nabble.com.

2 1

numpy.random selection from a loglogistic distribution?
by Liz VanWormer June 16, 2014

June 16, 2014

Dear all, I'm a novice Python user, and I need to draw random variables from a loglogistic distribution. I've used the numpy.random command in the past to select variables from supported distributions (beta, normal, lognormal, etc). Is there a way to add in a distribution to be used with the numpy.random command? Thank you for your insight, Liz

2 2