pyFFTW is an attempt to produce a pythonic wrapper around FFTW. The ultimate aim is to present a unified interface for all the possible transforms that FFTW can perform.
Both the complex DFT and the real DFT are supported, as well as on arbitrary axes of abitrary shaped and strided arrays, which makes it almost feature equivalent to standard and real FFT functions of numpy.fft (indeed, it supports the clongdouble dtype which numpy.fft does not). See the numpy.fft documentation for more information on that module.
The source can be found in github and it’s page in the python package index is here.
A comprehensive unittest suite is included with the source on the repository.
A quick (and very much non-comprehensive) usage example:
>>> import pyfftw, numpy
>>> # Create 3 16-byte aligned arays using the aligned array creation functions
>>> # They are cleared during the object creation, so there is no point filling them.
>>> a = pyfftw.n_byte_align_empty((1,4), 16, dtype=numpy.complex128)
>>> b = pyfftw.n_byte_align_empty(a.shape, 16, dtype=a.dtype)
>>> c = pyfftw.n_byte_align_empty(a.shape, 16, dtype=a.dtype)
>>> # Create the DFT and inverse DFT objects
>>> fft = pyfftw.FFTW(a, b)
>>> ifft = pyfftw.FFTW(b, c, direction='FFTW_BACKWARD')
>>> # Fill a with data
>>> a[:] = [1, 2, 3, 4]
>>> print a
[[ 1.+0.j 2.+0.j 3.+0.j 4.+0.j]]
>>> # perform the fft
>>> fft.execute()
>>> print b
[[ 10.+0.j -2.+2.j -2.+0.j -2.-2.j]]
>>> # perform the inverse fft
>>> ifft.execute()
>>> print c/a.size
[[ 1.+0.j 2.+0.j 3.+0.j 4.+0.j]]
Note that what was previously the ComplexFFTW class is now just called FFTW. Simply renaming the class should be sufficient to migrate