[Numpy-discussion] FFTS for numpy's FFTs (was: Re: Choosing between NumPy and SciPy functions)

Henry Gomersall heng at cantab.net
Tue Oct 28 05:46:59 EDT 2014

On 28/10/14 09:41, David Cournapeau wrote:
> The real issue with fftw (besides the license) is the need for plan 
> computation, which are expensive (but are not needed for each 
> transform). Handling this in a way that is user friendly while 
> tweakable for advanced users is not easy, and IMO more appropriate for 
> a separate package.

Just on this, I like to think I've largely solved the issue with:

If you have suggestions on how it can be improved, I'm all ears (there 
are a few things in the pipeline, like creating FFTW objects for 
different types of transform more explicit, which is likely to be the 
main difference for the next major version).



More information about the NumPy-Discussion mailing list