torch_heaviside: Heaviside

In torch: Tensors and Neural Networks with 'GPU' Acceleration

Heaviside

Description

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 documentation built on May 29, 2024, 9:54 a.m.