Log1p
Note
This function is more accurate than torch_log for small
values of input
log1p(input, out=NULL) -> Tensor
Returns a new tensor with the natural logarithm of (1 + input).
$$ y_i = \log_{e} (x_i + 1) $$
Examples
if (torch_is_installed()) {
a = torch_randn(c(5))
a
torch_log1p(a)
}
#> torch_tensor
#> 0.4720
#> 0.3297
#> 0.8045
#> nan
#> 0.9600
#> [ CPUFloatType{5} ]