[Numpy-discussion] Revised NEP-18, array_function protocol (original) (raw)

Matti Picus matti.picus at gmail.com
Thu Jun 28 20:34:32 EDT 2018

On 28/06/18 17:18, Stephan Hoyer wrote:

On Thu, Jun 28, 2018 at 1:12 PM Marten van Kerkwijk <m.h.vankerkwijk at gmail.com <mailto:m.h.vankerkwijk at gmail.com>> wrote:

For C classes like the ufuncs, it seems _self_ is defined for methods as well (at least, np.add.reduce._self_ gives np.add), but not a _func_. There is a _name_ (="reduce"), though, which means that I think one can still retrieve what is needed (obviously, this also means _arrayufunc_ could have been simpler...)

Good point! I guess this means we should encourage using name rather than func. I would not want to preclude refactoring classes from Python to C/Cython.

NumPy-Discussion mailing list NumPy-Discussion at python.org https://mail.python.org/mailman/listinfo/numpy-discussion There was opposition to that in a PR I made to provide a wrapper around matmul to turn it into a ufunc. It would have left the name but changed the func. https://github.com/numpy/numpy/pull/11061#issuecomment-387468084

More information about the NumPy-Discussion mailing list