rand_like

rand_like(
    x: &Tensor<T>, 
    low: T, 
    high: T
) -> Result<Tensor<T>, TensorError>

same as rand but the shape will be based on x.

Parameters:

x: input Tensor

low: the lowest value

high: the highest value

Returns:

Tensor with type T

Examples:

use hpt::{error::TensorError, ops::Random, Tensor};

fn main() -> Result<(), TensorError> {
    let a = Tensor::<f32>::rand([10, 10], 0.0, 10.0)?;
    println!("{}", a.rand_like(0.0, 10.0)?);
    Ok(())
}

Backend Support

Backend	Supported
CPU	✅
Cuda	✅