14 lines
826 B
Text
14 lines
826 B
Text
|
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.
|