Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Numba support for numpy.correlate()? #2500

Closed
echavess opened this issue Aug 10, 2017 · 4 comments
Closed

Numba support for numpy.correlate()? #2500

echavess opened this issue Aug 10, 2017 · 4 comments
Assignees
Milestone

Comments

@echavess
Copy link

Dear all,

I was trying to @jit(nopython=True) one of my functions which runs numpy.correlate() and found that Numba still does not support it. Is there any idea of when this feature is going to be supported? May I request, please, support to this function and others like numpy.convolve() ?? These two functions are very important for digital signal processing.

Thanks very much in advance

@echavess echavess changed the title Numba doesn't support numpy.correlate() Numba support for numpy.correlate()? Aug 10, 2017
@ThomasA
Copy link
Contributor

ThomasA commented Oct 5, 2017

I also ran into this issue just now.

@seibert seibert added this to the Numba 0.38 RC milestone Oct 31, 2017
@echavess
Copy link
Author

Hi guys, any updates? Will appreciate if Numba provides support to numpy.correlate and numpy.convolve.

Thanks in advance,

@stuartarchibald
Copy link
Contributor

@echavess I was working on this last week but had to go and do some other things. Hopefully it'll be done soon! Also, out of interest, is there a desire for FFT calls to be accepted by Numba, e.g. np.fft.fft ?

@stuartarchibald stuartarchibald self-assigned this Feb 20, 2018
@mbargull
Copy link
Contributor

Sure, that would be nice. numpy.correlate will just be too slow for larger inputs. Hence something like scipy.signal.fftconvolve, or more flexible, numpy.fft would be nice.
Some time ago I started to write some (incomplete/special-purpose) wrapper around MKL/FFTW via cffi to be used with Numba. In the meantime https://github.com/IntelPython/mkl_fft came out -- that might be of interest to people looking to implement fast convolution/correlation (I have yet to check that out...).

stuartarchibald added a commit to stuartarchibald/numba that referenced this issue Feb 28, 2018
As title. Lack of FFT support prevents the implementation of
efficient convolution alg.

Closes numba#2500
stuartarchibald added a commit to stuartarchibald/numba that referenced this issue Feb 28, 2018
As title. Lack of FFT support prevents the implementation of
efficient convolution alg.

Closes numba#2500
stuartarchibald added a commit to stuartarchibald/numba that referenced this issue Feb 28, 2018
As title. Lack of FFT support prevents the implementation of
efficient convolution alg.

Closes numba#2500
stuartarchibald added a commit to stuartarchibald/numba that referenced this issue Feb 28, 2018
As title. Lack of FFT support prevents the implementation of
efficient convolution alg.

Closes numba#2500
stuartarchibald added a commit to stuartarchibald/numba that referenced this issue Feb 28, 2018
As title. Lack of FFT support prevents the implementation of
efficient convolution alg.

Closes numba#2500
stuartarchibald added a commit to stuartarchibald/numba that referenced this issue Mar 1, 2018
As title. Lack of FFT support prevents the implementation of
efficient convolution alg.

Closes numba#2500
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

5 participants