juggernaut 0.9.0

Neural Network in Rust
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46

use activation::Activation;

#[derive(Copy, Clone)]
pub struct RectifiedLinearUnit;

impl RectifiedLinearUnit {
    pub fn new() -> RectifiedLinearUnit {
        return RectifiedLinearUnit;
    }
}

impl Activation for RectifiedLinearUnit {
    /// Calculates the RectifiedLinearUnit of input `x`
    fn calc(&self, x: Vec<f64>) -> Vec<f64> {
        x.iter()
            .map(|&n| if n <= 0f64 { 0f64 } else { n })
            .collect::<Vec<_>>()
    }

    /// Calculates the Derivative RectifiedLinearUnit of input `x`
    fn derivative(&self, x: Vec<f64>) -> Vec<f64> {
        x.iter()
            .map(|&n| if n <= 0f64 { 0f64 } else { n })
            .collect::<Vec<_>>()
    }
}

#[cfg(test)]
mod tests {
    use super::Activation;
    use super::RectifiedLinearUnit;

    #[test]
    fn rectifiedlinearunit_test() {
        let activation = RectifiedLinearUnit::new();
        assert_approx_eq!(activation.calc(vec![3.4f64])[0], 3.4f64);
        assert_approx_eq!(activation.calc(vec![-3.4f64])[0], 0f64);
    }

    #[test]
    fn rectifiedlinearunit_derivative_test() {
        let activation = RectifiedLinearUnit::new();
        assert_approx_eq!(activation.derivative(vec![-3.4f64])[0], 0f64);
        assert_approx_eq!(activation.derivative(vec![3.4f64])[0], 3.4f64);
    }
}