[SciPy-user] Failed build w/ mingw

Travis Brady travis.brady at gmail.com
Wed Jan 18 14:28:14 CST 2006


I went and deleted my Numpy and Scipy svn dirs and checked out from SVN and
tried again.
The build failed on the monte carlo stuff again (which was in the sandbox
setup.py), so I commented that bit out and it builds and tests fine.

Thanks for everybody's help.

Travis

On 1/17/06, Travis Oliphant <oliphant.travis at ieee.org> wrote:
>
> Travis Brady wrote:
>
> > Thanks, Travis.
> > The build succeeds after commenting out the montecarlo bit, but I
> > receive the following when trying to run the tests.
> >
> > >>> import scipy
> > import utils -> failed: No module named base
> > >>> scipy.test()
> > Overwriting lib=<module 'scipy.lib' from
> > 'C:\Python24\lib\site-packages\scipy\li
> > b\__init__.pyc'> from
> > C:\Python24\lib\site-packages\scipy\lib\__init__.pyc (was
> > <module 'numpy.lib ' from
> > 'C:\Python24\lib\site-packages\numpy\lib\__init__.pyc'>
> >  from C:\Python24\lib\site-packages\numpy\lib\__init__.pyc)
> > Fatal Python error: can't initialize module specfun (failed to import
> > scipy.base
> > )
>
>
> Hmm...  This is odd.  It looks like you are picking up an old header
> file from somewhere.  This failure is due to import_array() in the
> initialization and it's apparently trying to load scipy.base.  This used
> to be what was loaded to get the C-API.  But, now it's numpy.core.
>
> Make sure you are not picking up old versions of the header when you
> compile scipy...
>
> -Travis
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.net
> http://www.scipy.net/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.scipy.net/pipermail/scipy-user/attachments/20060118/a181a328/attachment.htm


More information about the SciPy-user mailing list