BUG: `full_like` accepts array `fill_value`

Question

lucascolley opened this issue 4 months ago · 3 comments

In [1]: import array_api_strict as xp

In [2]: xp.full_like(xp.asarray(0), xp.asarray(1))
Out[2]: Array(1, dtype=array_api_strict.int64)

fill_value (Union[bool, int, float, complex]) – fill value.

Answer 1 · 2024-08-08T18:50:03.000Z

This seems like it should be fixed in the standard. Getting a fill value from an array seems like it wouldn't be uncommon.

Answer 2 · 2024-08-08T20:08:53.000Z

Torch only accepts a scalar in full and full_like. To update the standard, I'd want to see upstream support added first.

Answer 3 · 2024-08-08T20:17:25.000Z

Dask also throws a FutureWarning since its implementation uses np.copyto, which it says may stop working with dask arrays in the future.