issues: 2021810083

This data as json

id node_id number title user state locked assignee milestone comments created_at updated_at closed_at author_association active_lock_reason draft pull_request body reactions performed_via_github_app state_reason repo type
2021810083 PR_kwDOAMm_X85g8r6c 8508 Implement `np.clip` as `__array_function__` 5635139 closed 0     2 2023-12-02T02:20:11Z 2023-12-03T05:27:38Z 2023-12-03T05:27:33Z MEMBER   0 pydata/xarray/pulls/8508

Would close https://github.com/pydata/xarray/issues/2570

Because of https://numpy.org/neps/nep-0018-array-function-protocol.html#partial-implementation-of-numpy-s-api, no option is ideal: - Don't do anything — don't implement __array_function__. Any numpy function that's not a ufunc — such as np.clip will materialize the array into memory. - Implement __array_function__ and lose the ability to call any non-ufunc-numpy-func that we don't explicitly configure here. So np.lexsort(da) wouldn't work, for example; and users would have to run np.lexsort(da.values). - Implement __array_function__, and attempt to handle the functions we don't explicitly configure by coercing to numpy arrays. This requires writing code to walk a tree of objects looking for arrays to coerce. It seems to go against the original numpy proposal.

@shoyer is this summary accurate?

 
{
    "url": "https://api.github.com/repos/pydata/xarray/issues/8508/reactions",
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
     13221727 pull

Links from other tables