lucid.nonzero¶

lucid.nonzero(a: Tensor, /) → Tensor¶

The nonzero function returns the indices of non-zero elements in the input tensor. It outputs a 2D tensor where each row corresponds to the index of a non-zero element.

Function Signature¶

def nonzero(a: Tensor) -> Tensor

Parameters¶

a (Tensor): The input tensor from which to find non-zero indices.

Returns¶

Tensor: A 2D tensor of shape \((N, D)\) where \(N\) is the number of non-zero elements and \(D\) is the number of dimensions in the input tensor.

Each row of the output represents a coordinate index of a non-zero element.

The output tensor has int32 dtype and does not track gradients.

Note

This operation is non-differentiable. Gradients will not be propagated through nonzero().

Examples¶

>>> import lucid
>>> x = lucid.tensor([[1, 0], [0, 2]])
>>> lucid.nonzero(x)
Tensor([[0, 0],
        [1, 1]], grad=None, device=cpu)

>>> x = lucid.tensor([0, 3, 0, 4])
>>> lucid.nonzero(x)
Tensor([[1],
        [3]], grad=None, device=cpu)