Skip to contents

Heaviside

Usage

torch_heaviside(self, values)

Arguments

self

(Tensor) the input tensor.

values

(Tensor) The values to use where input is zero.

heaviside(input, values, *, out=None) -> Tensor

Computes the Heaviside step function for each element in input. The Heaviside step function is defined as:

$$ \mbox{{heaviside}}(input, values) = \begin{array}{ll} 0, & \mbox{if input < 0}\\ values, & \mbox{if input == 0}\\ 1, & \mbox{if input > 0} \end{array} $$

Examples

if (torch_is_installed()) {

input <- torch_tensor(c(-1.5, 0, 2.0))
values <- torch_tensor(c(0.5))
torch_heaviside(input, values)
values <- torch_tensor(c(1.2, -2.0, 3.5))
torch_heaviside(input, values)
}
#> torch_tensor
#>  0
#> -2
#>  1
#> [ CPUFloatType{3} ]