[Numpy-discussion] GSoC'15 - mentors & ideas

Sebastian Berg sebastian at sipsolutions.net
Thu Feb 26 10:25:06 EST 2015


On Do, 2015-02-26 at 07:09 -0800, Jaime Fernández del Río wrote:
<snip>
>         
>         
> To add one of my own: the old iterator is still being used in many,
> many places throughout the numpy code base. Wouldn't it make sense to
> port those to the new one? In doing so, it would probably lead to
> producing simplified interfaces to the new iterator, e.g. reproducing
> the old PyIter_AllButAxis is infinitely more verbose with the new
> iterator.
> 
> 

Might be a bit off topic. But I used to wonder if it could make sense to
create a Cython code generation support for nditer? NDiter is pretty
powerful, but we often have things like the contiguous special case,
buffering, etc. that is always identical code but without having
something ready in cython nobody will ever use nditer from cython, even
though for some things it might make a lot of sense.

- Sebastian

<snap>
> 
> _______________________________________________
> NumPy-Discussion mailing list
> NumPy-Discussion at scipy.org
> http://mail.scipy.org/mailman/listinfo/numpy-discussion

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: This is a digitally signed message part
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20150226/8e1126ec/attachment.sig>


More information about the NumPy-Discussion mailing list