Boolean mask + ellipsis lead to incorrect indexing?

1

Indexing with boolean masks and ellipsis leads to inconsistent behavior.

MWE:

import torch
t = torch.arange(12).view(3,2,-1)
mask = torch.BoolTensor([[True, False], [False, True], [False, False]])

the expression

t[mask][...,0]

returns

>>> tensor([0, 6])

whereas

t[mask,..., 0]

returns

>>> tensor([0, 5])

With numpy arrays, both [mask,..., 0] and [mask][..., 0] return >>> array([0, 6]).

2

This seems indeed unexpected. Could you create an issue on GitHub so that we could track and fix it, please?

3

github issue submitted: boolean mask + ellipsis lead to incorrect indexing · Issue #70347 · pytorch/pytorch · GitHub