polyfit-rs 0.2.2

Polynomial fitting of data like in numpy.polyfit().
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

extern crate rand;

#[cfg(test)]
mod tests {
    use rand::{Rng, SeedableRng};
    use polyfit_rs as poly;

    #[test]
    fn simple_parabola_fitting() {
        // f(x) = a*x² + b * x + c
        // c = 1.0, b = 1.1, a = 2.0
        // Order of monomials increases with the vector index
        let parameters = vec![1.0, 1.1, 2.0];
        let mut y_values: Vec<f64> = Vec::with_capacity(210);
        let mut y_values_with_errors: Vec<f64> = Vec::with_capacity(210);
        let mut x_values: Vec<f64> = Vec::with_capacity(210);

        let mut rng = rand::rngs::StdRng::seed_from_u64(5000u64);
        let epsilon = 10.0f64;

        for i in -100..110
        {
            let x = (i as f64) / 10.0;
            let y: f64 = parameters[0] + parameters[1] * x + parameters[2] * x * x;
            let error: f64 = rng.random_range(-epsilon..epsilon);
            let y_with_error = y + error;
            y_values_with_errors.push(y_with_error);
            x_values.push(x);
            y_values.push(y);
        }

        let fitted_parameters = poly::polyfit_rs::polyfit(&x_values, &y_values_with_errors, 2).unwrap();
        
        let mut i = 0;
        for x in x_values
        {
            let fitted_y: f64 = fitted_parameters[0] + fitted_parameters[1] * x + fitted_parameters[2] * x * x;
            let error = (fitted_y - y_values[i]).abs();
            assert!(error <= epsilon);
            i += 1;
        }
    }
}