Repository navigation
Provide Dask interface to mkl_fft #5
Description
Activity
Any thoughts on this, @oleksandr-pavlyk?
@jakirkham I think dask interface is a great idea. I am not clear as to what exactly should be done in
_numpy_fft.Are you proposing to add
_dask_fft.pyto the package, which would export thesefft_wrapped function ifdaskis found in the environment?Thanks for the follow up. Glad to hear this works.
I am not clear as to what exactly should be done in
_numpy_fft.Don’t think anything has to be done to
_numpy_fft. That just works well for this purpose. Namelyfft_wrapworks well with FFT functions adhering to the NumPy API.Are you proposing to add
_dask_fft.pyto the package, which would export thesefft_wrapped function ifdaskis found in the environment?Yep, exactly.
Any thoughts on how we name things in the public API? Also any thoughts on what would be adequate testing?
Would be nice if a simple
daskinterface were provided tomkl_fft(much like thenumpyone currently). In fact it could be built right on top of the_numpy_fftmodule by wrapping these functions usingfft_wrapjust as it is done in Dask except using the one's in_numpy_fftinstead.