[SciPy-user] fftw, scipy ("old") install

[Arnd Baecker]

Hi Pearu,

On Mon, 17 Oct 2005, Pearu Peterson wrote:

> On Mon, 17 Oct 2005, Arnd Baecker wrote:

> > it seems that I failed with fftw on the opteron ;-) -
> > scipy fft is (much) slower than the Numeric one
> > (see output below).
> >
> > I installed fftw via::
> >
> >  tar xzf ../Sources/fftw-3.0.1.tar.gz
> >  cd fftw-3.0.1/
> >  ./configure CFLAGS=-fPIC  --prefix=$PHOME
> >  make
> >  make install
>
> The old scipy.fftpack supports only fftw2. I have a patch for fftw3
> somewhere. I'll apply it as soon as I'll get a chance to work on fftpack.

I thought it should work - at least that's how I understood
your message
  http://aspn.activestate.com/ASPN/Mail/Message/scipy-user/2811157

So I went for
  svn co http://svn.scipy.org/svn/scipy/trunk scipy
  cd scipy
  svn co http://svn.scipy.org/svn/scipy_core/trunk/ scipy_core

which gives:
  In [1]: import scipy
  In [2]: scipy.__version__
  Out[2]: '0.3.3_303.4573'

Was this the wrong branch etc.?


> > Reconsidering this after seeing the poor performance
> > shows that, `nm _fftpack.so` gives me no symbols  related to fftw.
>
> Yes, _fftpack.so should contain Fortran fftpack when fftw (v2) library is
> not available.
>
> > And for future builds:
> > Is there a way to check (on the python level) which fft is used
> > *after* the installation?
>
> Normally it should be unnecessary.  However, scipy building scripts install
> config.py (old scipy) or __config__.py (newscipy) files containing the
> system information used in building scipy.

I do think that this is important (so one can check,
which capabilities are really there.)
So thanks a lot for pointing out config.py - this contains
what I was looking for!

Best, Arnd