[SciPy-Dev] Scipy Central Home Page .. finally

Pauli Virtanen pav@iki...
Sun Feb 17 07:10:52 CST 2013


Hi,

17.02.2013 07:03, Surya Kasturi kirjoitti:
[clip]
> So, I am even interested in writing some Python.. (if there is anything
> specific I can try.. small initially..).

Some suggestions given previously by Ralf:

http://thread.gmane.org/gmane.comp.python.scientific.devel/17173

This one should be quick to do (< 1 hour of work):

http://projects.scipy.org/scipy/ticket/1621

The fftconvolve() routine in scipy.signal currently pads arrays to
powers of two. However, the FFTPACK engine works well (I think) also
with factors of 2, 3, and 5. (You might have to do some benchmarking or
looking into the FFTPACK fortran source files to verify this.)

That ticket already contains a function to do this (for some other sets
of factors), so the only thing here is to integrate this to
scipy.signal. If you want some practice, taking this, and forming it up
to a pull request in Scipy could be a useful idea.

-- 
Pauli Virtanen



More information about the SciPy-Dev mailing list