[SciPy-dev] FFTPACK and RandomArray

eric eric at scipy.org
Tue Feb 19 22:47:54 CST 2002

Just to fill everyone in, we're gonna be making fftw an "optional" package in
SciPy.  This is because its license doesn't fit with the rest of the package.
FFTPACK provides pretty much a drop in replacement in functionality, so no
capabilities are lost.  The only drawpack is that fftw is faster.  On the
upside, it will simplify the build process, and the part of the pain that
Fernando suffered today will be alleviated.

> I like it.
> How should we handle fftpack.  Should we use f2py and modified Numeric
> wrappers?  Or use the f2c version Numeric currently uses.

Seems the fastest (easiest to get working) solution is just to use the one that
comes with Numeric.  You probably loose some speed though in the f2c conversion
process.  I don't know how many functions there are to wrap, but making f2py
wrappers shouldn't be that difficult.  This should be the "long term goal" I
think -- it can even be the short term goal.  Who ever does it gets to decide.

> Also, I've been working on the stats package.  How should we handle
> RandomArray.  Should we take all that's good from RandomArray and place it in
> our stats package?


By the way.  How is this coming?  Are you going the multivariate or
singlevariate route -- or has that been decided yet?

> My inclination is to subsume RandomArray, FFTPACK, and LinearAlgebra into
> scipy so that all we rely on is the core array facilities of Numeric.



More information about the Scipy-dev mailing list