Hi, Is there any reason to have separate functions - or to keep enforcing that? I agree, an equivalent of R's rm.na argument seems like a very reasonable and useful addition, such as (sorry for the obviousness):
np.mean(x, dropna=True) and so on, Cheers, Matthew On Mon, Oct 9, 2023 at 9:17 AM Juan Nunez-Iglesias <[email protected]> wrote: > > > On Mon, 9 Oct 2023, at 7:07 PM, Andrew Nelson wrote: > > On Mon, 9 Oct 2023 at 16:36, Jerome Kieffer <[email protected]> wrote: > I'd be ambivalent on making this change. THere are a whole host of other > `np.nan*` functions, would they all need to be modified as well? e.g. > nanprod, nansum, nanargmin, ...... > > > I think obviously, either change all functions or none. The question is > whether such a change would fit into the overall NumPy 2.0 and array-API > plans. 🤷♂️ > _______________________________________________ > NumPy-Discussion mailing list -- [email protected] > To unsubscribe send an email to [email protected] > https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ > Member address: [email protected] _______________________________________________ NumPy-Discussion mailing list -- [email protected] To unsubscribe send an email to [email protected] https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: [email protected]
