Update on "torch.sgn for complex tensors"

anjali411 · anjali411 · commit 9354172940cd · 2020-08-04T13:33:13.000-07:00
resolves #36323 by adding `torch.sgn` for complex tensors. `torch.sgn` returns `x/abs(x)` for `x != 0` and returns `0 + 0j` for `x==0` This PR doesn't test the correctness of the gradients. It will be done as a part of auditing all the ops in future once we decide the autograd behavior (JAX vs TF) and add gradchek. [ghstack-poisoned]
diff --git a/torch/_torch_docs.py b/torch/_torch_docs.py
@@ -5725,7 +5725,7 @@ def merge_dicts(*dicts):
 
 add_docstr(torch.sgn,
            r"""
-sgn(input, out=None) -> Tensor
+sgn(input, *, out=None) -> Tensor
 
 Returns a new complex tensor with the same angle as that of the elements of :attr:`input` and absolute value 1.
 
