.is_nonzero()

In PyTorch, the .is_nonzero() function is used to check if a scalar tensor contains a non-zero element. It returns True if the tensor’s element is non-zero and False if it is zero. This method is particularly useful for conditional checks or validation in tensor operations.

Syntax

torch.is_nonzero(input)

• input: Tensor with a single element.

Example

This code demonstrates how PyTorch’s .is_nonzero() function checks whether a tensor contains a non-zero value:

import torch

# Define a tensor with a non-zero element
tensor1 = torch.tensor(5.0)

# Check if the tensor contains a non-zero element
result1 = torch.is_nonzero(tensor1)
print(result1)

# Define a tensor with a zero element
tensor2 = torch.tensor(0.0)

# Check if the tensor contains a non-zero element
result2 = torch.is_nonzero(tensor2)
print(result2)

The above code produces the following output:

True
False