Rate this Page
★★★★★

torch.masked_select#

torch.masked_select(input,mask,*,out=None)→Tensor#

Returns a new 1-D tensor which indexes theinput tensor according tothe boolean maskmask which is aBoolTensor.

The shapes of themask tensor and theinput tensor don’t needto match, but they must bebroadcastable.

Note

The returned tensor doesnot use the same storageas the original tensor

Parameters

  • input (Tensor) – the input tensor.

  • mask (BoolTensor) – the tensor containing the binary mask to index with

Keyword Arguments

out (Tensor,optional) – the output tensor.

Example:

>>>x=torch.randn(3,4)>>>xtensor([[ 0.3552, -2.3825, -0.8297,  0.3477],        [-1.2035,  1.2252,  0.5002,  0.6248],        [ 0.1307, -2.0608,  0.1244,  2.0139]])>>>mask=x.ge(0.5)>>>masktensor([[False, False, False, False],        [False, True, True, True],        [False, False, False, True]])>>>torch.masked_select(x,mask)tensor([ 1.2252,  0.5002,  0.6248,  2.0139])