[Numpy-discussion] PyArray_GETITEM and PyArray_SETITEM

Eric Wieser wieser.eric+numpy at gmail.com
Fri Nov 17 16:44:29 EST 2017


It’s worth noting that PyArray_GETITEM is the equivalent of arr[....].item(),
not arr[...]. If you want the behavior of the latter, use PyArray_Scalar
instead.

Similarly, PyArray_SETITEM is only guaranteed to be equivalent to arr[...]
= x when isinstance(x, np.generic) is false.

I don’t think these belong in public API yet, because they don’t expose the
interface that most people might expect. Their names are based solely on
the names of descr->f->getitem.

Eric
​

On Fri, 17 Nov 2017 at 13:12 Allan Haldane <allanhaldane at gmail.com> wrote:

> On 11/13/2017 01:53 PM, Mmanu Chaturvedi wrote:
> > Hello All,
> >
> > I need to make use of the limited numpy API access Pybind11 gives, in
> order
> > to add a feature to it.  It seems to give access to functions from
> > numpy_api.py [1].  I need to use PyArray_GETITEM and PyArray_SETITEM in
> > order to get and set array elements [2], these functions / macros  are
> not
> > exposed via numpy_api.py, but are in `numpy/ndarraytypes.h`.
> >
> > We were wondering why aren't PyArray_GETITEM and PyArray_SETITEM exposed
> > like the rest of numpy API?  Is it possible to replicate the behavior
> using
> > the members exposed in numpy_api.py ?  Any help would be appreciated.
> >
> > Mmanu
>
> It looks like that was the plan. There are comments there saying they
> would become part of the API in "numpy 2.0" (which hasn't happened yet).
>
> In the meantime, maybe you can use PySequence_SetItem? I expect that
> there is only very minimal overhead in using that vs PyArray_SETITEM.
>
> Allan
> _______________________________________________
> NumPy-Discussion mailing list
> NumPy-Discussion at python.org
> https://mail.python.org/mailman/listinfo/numpy-discussion
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20171117/c12b1344/attachment-0001.html>


More information about the NumPy-Discussion mailing list