ENH: add a naive divmod, un-xfail relevant tests #84
Conversation
|
Two more ufuncs with two optional out= inputs and two OutArray outputs : ldexp and frexp. Need a dedicated result annotation, apparently. |
ev-br
force-pushed
the
return_annotation
branch
from
35e647b to
bc9bd0c
Compare
| quot = floor_divide(x1, x2, out=out1, **kwds) | ||
| rem = remainder(x1, x2, out=out2, **kwds) |
There was a problem hiding this comment.
Note that here (and in general) if we want these functions to be differentiable in PyTorch, we should not use their out= variant. We should implement the out= behaviour manually.
There was a problem hiding this comment.
Here these are wrapper floor_divide and remainder, so torch.remainder and torch.floor_divide never see the out=.
Also NB: this will need a rework anyway, apparently there are more (out1, out2) ufuncs, ldexp and frexp. So Tuple[NDArray] return annotation will need to appear after we settle on the generic machinery.
There was a problem hiding this comment.
Right! And sure, let's just push this discussion then. I still think that the out= machinery can be implemented in a generic way.
| # postprocess return values | ||
|
|
||
|
|
||
| def postprocess_ndarray(result, **kwds): | ||
| return _helpers.array_from(result) | ||
|
|
||
|
|
||
| def postprocess_out(result, **kwds): | ||
| result, out = result | ||
| return _helpers.result_or_out(result, out, **kwds) | ||
|
|
||
|
|
||
| def postprocess_tuple(result, **kwds): | ||
| return _helpers.tuple_arrays_from(result) | ||
|
|
||
|
|
||
| def postprocess_list(result, **kwds): | ||
| return list(_helpers.tuple_arrays_from(result)) | ||
|
|
||
|
|
||
| def postprocess_variadic(result, **kwds): | ||
| # a variadic return: a single NDArray or tuple/list of NDArrays, e.g. atleast_1d | ||
| if isinstance(result, (tuple, list)): | ||
| seq = type(result) | ||
| return seq(_helpers.tuple_arrays_from(result)) | ||
| else: | ||
| return _helpers.array_from(result) | ||
|
|
||
|
|
||
| postprocessors = { | ||
| NDArray: postprocess_ndarray, | ||
| OutArray: postprocess_out, | ||
| NDArrayOrSequence: postprocess_variadic, | ||
| tuple[NDArray]: postprocess_tuple, | ||
| list[NDArray]: postprocess_list, | ||
| } | ||
|
|
||
|
|
There was a problem hiding this comment.
This is duplicated from the other PR.
| postprocess = postprocessors.get(r, None) | ||
| if postprocess: | ||
| kwds = {"promote_scalar": promote_scalar_out} | ||
| result = postprocess(result, **kwds) | ||
| return result |
That's just an artefact from giving optional return arguments to |
|
Merged in #91 |
Implement
np.divmod(x1, x2)andndarray.__divmod__. These are unusual in that they allow two output arrays, so it's either complicate and obfuscate the ufunc machinery just for these two, or special case. This PR opts for the latter :-).Since there is no pytorch-native divmod (cf pytorch/pytorch#90820 ), this PR just does
(x1 // x2, x1 % x2)in the wrapper level.This PR is on top of gh-83, for clarity.