[SciPy-user] Fourier-sine transform with SciPy

Lubos Vrbka lists@vrbka....
Fri Jun 6 04:17:44 CDT 2008


thanks for quick reply!

> I don't think we have a function that computes discrete sine or cosine
> transforms, but if f is real, you can get the sine transform you wrote
> as the imaginary part of a complex Fourier transform.
by this you mean using scipy.fftpack.rfft()? i must admit i am a bit 
lost in the various possibilities that are available.

scipy.fftpack.*fft* functions are clear (is this the netlib implementation?)

scipy.ifft?? tells me (among others)
File:             /usr/lib/python2.4/site-packages/numpy/fft/fftpack.py

or should i use numpy.fft? what's the difference in fftpack in scipy and 

is there any place where these things are documented, i.e. what code 
does actually lie behind these functions?

and just out of curiosity - from what i've seen during my search for the 
  information on google, fftw was (at least in the past) used in scipy. 
the scipy package on my debian box depends on fftw2, but the names refer 
to fftpack, that shouldn't be fftw but netlib implementation instead (if 
i am not wrong). could anybody clarify this for me, please? sorry if 
this was discussed before, but i wasn't able to find the relevant thread.

thanks once more!

Lubos _@_"

More information about the SciPy-user mailing list