[SciPy-user] Issues building from source on windows (SVN)

[Ryan Krauss]

I am going to give my students an executable that is built for P3's
and AMD's if the one from scipy.org for SSE2 chips doesn't work for
them.  I mainly followed the instructions from here
http://scipy.org/Installing_SciPy/Windows
ignoring the section about Intel MKL stuff.  So basically I did 5 things:
1. Check out Numpy and Scipy from SVN (download tortoisesvn
http://tortoisesvn.tigris.org/)
2. Setup a basic MinGW compiler system (I find this to be more painful
than the MinGW people seem to think it is.  You need to download
almost all of the bin packages they provide and unpack them into a
folder on your path so that you can type "gcc -v" in a cmd.exe window
and get some sensible response.  You need to make sure you get
gcc-core, gcc-g77, mingw-runtime, and w32-api.  The mingw-runtime
contains things like stdio.h.)
3. Download the precomiled ATLAS binaries (follow link on
Installing_Scipy/Windows parge)
4. Create a site.cfg file that contains
[atlas]
library_dirs = c:\path\to\BlasLapackLibs
atlas_libs = lapack, f77blas, cblas, atlas
5. type c:\path\to\python.exe setup.py config --compiler=mingw32 build
--compiler=mingw32 bdist_wininst
at a cmd.exe prompt.

The exe file created isn't perfect, there are some test failures.  But
I think it meets my needs.  If your email can handle 2 meg
attachments, I can send you my exe file to try.  I don't know if you
need the numpy and the scipy one or if my scipy will work with your
numpy.

Ryan

On 5/22/07, Giulio Venezian <gvenezian at yahoo.com> wrote:
> excuse me if i keep barging into this conversation.
>
> Ryan: Do I gather from all this that you have
> succeeded in constructing an AMD-compatible version of
> SciPy?
>
> Would you please share what you've learned? As I
> mentioned before I'm trying to get the functions
> special.jn_zeros and special.jnp_zeros to work on my
> AMD-based machine. Other than using the
> self-installing version, I don't understand how one
> goes about constructing a version of SciPy.
>
> I'd appreciate it very much if you would email me
> whatever instructions you prepare for your students.
>
> Giulio
> --- Robert Kern <robert.kern at gmail.com> wrote:
>
> > Ryan Krauss wrote:
> > > Is there an easy way to test every package except
> > ndimage?  Because it
> > > fails and completely crashes Python, I don't know
> > any tests that run
> > > after ndimage also fail.
> >
> > from numpy import NumpyTest
> >
> > packages = """
> > scipy.cluster
> > scipy.fftpack
> > scipy.integrate
> > scipy.interpolate
> > scipy.io
> > scipy.lib
> > scipy.linalg
> > scipy.linsolve
> > scipy.maxentropy
> > scipy.misc
> > scipy.odr
> > scipy.optimize
> > scipy.signal
> > scipy.sparse
> > scipy.special
> > scipy.stats
> > scipy.stsci
> > scipy.weave
> > """.strip().split()
> >
> > for subpkg in packages:
> >     print subpkg
> >     t = NumpyTest(subpkg)
> >     t.test(1, 2)
> >
> > --
> > 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
> >
>
>
>
>
> ____________________________________________________________________________________
> Sucker-punch spam with award-winning protection.
> Try the free Yahoo! Mail Beta.
> http://advision.webevents.yahoo.com/mailbeta/features_spam.html
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>