ANFFT is an FFT package for Python, based on FFTW. It provides a
multi-threaded, self-tuning FFT interface via high-level functions
similar to the "fft" and "fftn" routines found in NumPy and SciPy.
ANFFT is intended to be used in situations where large numbers of expensive
FFTs are required, for which the built-in NumPy or SciPy FFTs are not
acceptable. By default, ANFFT provides immediate results by using
FFTW's "estimate" mode, which does not require tuning. However, each
high-level function provides a keyword named "measure" which will invoke the
full FFTW planning machinery. Plans for a given shape and type of array
are cached for the length of a Python session, and accummulated FFTW "wisdom"
is stored across Python sessions in a dotfile. You don't need to know anything
about FFTW internals to use ANFFT.
b978cdd4b5