anofox-ml-regression 0.1.0

Classical regression models (OLS, Ridge, Lasso, Elastic Net, GLM, GP) for anofox-ml
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

//! Conversion helpers between ndarray and faer types.

use faer::{Col, Mat};
use ndarray::{Array1, Array2};

/// Convert an ndarray `Array2<f64>` to a faer `Mat<f64>`.
pub fn ndarray_to_mat(x: &Array2<f64>) -> Mat<f64> {
    let (nrows, ncols) = x.dim();
    Mat::from_fn(nrows, ncols, |i, j| x[[i, j]])
}

/// Convert an ndarray `Array1<f64>` to a faer `Col<f64>`.
pub fn ndarray_to_col(y: &Array1<f64>) -> Col<f64> {
    Col::from_fn(y.len(), |i| y[i])
}

/// Convert a faer `Col<f64>` to an ndarray `Array1<f64>`.
pub fn col_to_ndarray(c: &Col<f64>) -> Array1<f64> {
    Array1::from_vec((0..c.nrows()).map(|i| c[i]).collect())
}

#[cfg(test)]
mod tests {
    use super::*;
    use approx::assert_abs_diff_eq;
    use ndarray::{array, Array2};

    #[test]
    fn test_roundtrip_mat() {
        let x = Array2::from_shape_vec((3, 2), vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0]).unwrap();
        let mat = ndarray_to_mat(&x);
        assert_eq!(mat.nrows(), 3);
        assert_eq!(mat.ncols(), 2);
        assert_abs_diff_eq!(mat[(0, 0)], 1.0);
        assert_abs_diff_eq!(mat[(2, 1)], 6.0);
    }

    #[test]
    fn test_roundtrip_col() {
        let y = array![1.0, 2.0, 3.0];
        let col = ndarray_to_col(&y);
        let back = col_to_ndarray(&col);
        assert_abs_diff_eq!(back[0], 1.0);
        assert_abs_diff_eq!(back[2], 3.0);
    }
}