Evgeni Burovski
Evgeni Burovski
Would be great to update the issue title/test then, if we repurpose it for `__setitem__` specifically.
> All real frameworks succeed, but array-api-strict fails because the standard is not defined for this case. If indeed _all real frameworks succeed_, then it very well could be added...
> We should keep our eyes out for any other situations where we need to skip for array-api-strict and reassess In SciPy, it'd be great to keep an explicit list...
So it might be worth revisiting this in the spec then.
Assuming that `List[array]` is the correct annotation for functions which return lists of arrays, the only functions which return a list of arrays are `meshgrid` and `broadcast_arrays`, cf https://github.com/data-apis/array-api/issues/934. Two...
@jakevdp while not exactly the _master list_ , the functions which return `List[array]` in the Array API spec but have been changed in NumPy 2.0 to return tuples are `meshgrid`...
Yes, numpy's concatenate is limited by what is in this guidance. Numpy, however, has np.r_, which --- with all is sins --- allows extending arrays with scalars or 0D arrays....
> np.concat is not limited by this guidance, it was always limited (check np.concatenate in older versions) This is exactly what I am saying: `np.concatenate` was always limited, and `np.r_`...
Exactly. I've a scipy helper, and this issue is to gauge interest/possibility to make it work with `xp.concat` (or `xp.stack`) and drop the helper.
This runs into https://github.com/data-apis/array-api/issues/923 on `jax.numpy`.