[Numpy-discussion] np.dot and scipy sparse matrices

Stéfan van der Walt stefan@sun.ac...
Wed Nov 9 13:27:11 CST 2011


Hi Peter

On Wed, Nov 9, 2011 at 3:38 AM, Peter Prettenhofer
<peter.prettenhofer@gmail.com> wrote:
> I recently got the latest numpy version (2.0.0.dev-7297785) from the
> git repo and realized that `np.dot` causes a segfault if its operands
> are scipy sparse matrices. Here's some code to reproduce the problem::
>
>   import numpy as np
>   from scipy import sparse as sp
>   A = np.random.rand(10, 10)
>   S = sp.csr_matrix(A)
>   _ = np.dot(A, A)   # this works OK
>   _ = np.dot(S, S)   # this segfaults!

Here's a patch to get rid of the segfault:

https://github.com/numpy/numpy/pull/170

I don't know if np.dot is supposed to know anything about sparse
matrices; if so, then a fancier change may be needed.

Regards
Stéfan


More information about the NumPy-Discussion mailing list