[Numpy-discussion] Numpy 1.10.2rc1

Charles R Harris charlesr.harris at gmail.com
Fri Nov 20 16:37:33 EST 2015


On Fri, Nov 20, 2015 at 2:00 PM, Charles R Harris <charlesr.harris at gmail.com
> wrote:

>
>
> On Fri, Nov 20, 2015 at 1:42 PM, Orion Poplawski <orion at cora.nwra.com>
> wrote:
>
>> On 11/12/2015 02:11 PM, Charles R Harris wrote:
>> > Hi All,
>> >
>> > I am pleased to announce the release of Numpy 1.10.2rc1. This release
>> should
>> > fix the problems exposed in 1.10.1, which is not to say there are no
>> remaining
>> > problems. Please test this thoroughly, exspecially if you experienced
>> problems
>> > with 1.10.1. Julian Taylor has opened an issue relating to cblas
>> detection on
>> > Debian (and probably Debian derived distributions) that is not dealt
>> with in
>> > this release. Hopefully a solution will be available before the final.
>>
>> So, this fails:
>>
>>   File "setup.py", line 427, in fortran_extensionlists
>>     if StrictVersion(np.version.version) > StrictVersion("1.6.1"):
>>   File "/usr/lib64/python2.7/distutils/version.py", line 40, in __init__
>>     self.parse(vstring)
>>   File "/usr/lib64/python2.7/distutils/version.py", line 107, in parse
>>     raise ValueError, "invalid version number '%s'" % vstring
>> ValueError: invalid version number '1.10.2rc1'
>>
>> But I'm not sure numpy has made any contracts to follow the distutils
>> StrictVersion format:
>>
>> http://epydoc.sourceforge.net/stdlib/distutils.version.StrictVersion-class.html
>
>
> No, we don't support StrictVersion nor does Scipy.
>
> """Utility to compare (Numpy) version strings.
>
> The NumpyVersion class allows properly comparing numpy version strings.
> The LooseVersion and StrictVersion classes that distutils provides don't
> work; they don't recognize anything like alpha/beta/rc/dev versions.
>
> """
>
> Looks like `numpy/distutils/mingw32ccompiler.py` needs fixing. Could you
> open an issue? The import of StrictVersion dates back to 2005, so not sure
> why this is turning up now. Maybe it is specific to compiling Fortran and
> we haven't done that with rc's before.
>

In fact, I don't see where the call is coming from. Is this something
specific to your project? If so, you want to use NumpyVersion which you can
import from `numpy.lib`.

Chuck
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20151120/712f4a09/attachment.html>


More information about the NumPy-Discussion mailing list