jiro_nn 0.8.1

Neural Networks framework with model building & data preprocessing features.
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
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166

#[macro_use]
extern crate assert_float_eq;

use jiro_nn::linalg::{Matrix, MatrixTrait, Scalar};

#[test]
fn test_zeros() {
    let m = Matrix::zeros(2, 3);
    for row in m.get_data_row_leading() {
        for val in row {
            assert_float_relative_eq!(val, 0.0, 0.00001);
        }
    }
}

#[test]
fn test_random_uniform() {
    let mat = Matrix::random_uniform(2, 2, 0.0, 1.0);
    assert_eq!(mat.dim().0, 2);
    assert_eq!(mat.dim().1, 2);
    assert!(mat
        .get_data_row_leading()
        .iter()
        .all(|row| { row.iter().all(|&val| val >= 0.0 && val < 1.0) }));
}

#[test]
fn test_random_normal() {
    let mat = Matrix::random_normal(100, 100, 0.0, (1.0 as Scalar).sqrt());
    assert_eq!(mat.dim().0, 100);
    assert_eq!(mat.dim().1, 100);
    // Check that the values are roughly normally distributed
    let mean = mat.get_data_row_leading().iter().flatten().sum::<Scalar>() / 10000.0;
    let variance = mat
        .get_data_row_leading()
        .iter()
        .flatten()
        .map(|val| (val - mean).powi(2))
        .sum::<Scalar>()
        / 1000.0;

    assert!(mean.abs() < 0.1);
    assert!((variance.abs() - 1.0).abs() < 0.1);
}

#[test]
fn test_from_iter() {
    let data = [1.0, 2.0, 3.0, 4.0, 5.0, 6.0];
    let m = Matrix::from_iter(2, 3, data.iter().cloned());
    assert_float_relative_eq!(m.get_data_col_leading()[0][0], 1.0, 0.00001);
    assert_float_relative_eq!(m.get_data_col_leading()[2][1], 6.0, 0.00001);
}

#[test]
fn test_from_row_leading_vector2() {
    let m = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    assert_float_relative_eq!(m.get_data_col_leading()[0][0], 1.0, 0.00001);
    assert_float_relative_eq!(m.get_data_col_leading()[2][1], 6.0, 0.00001);
}

#[test]
fn test_from_column_vector() {
    let m = Matrix::from_column_vector(&vec![1.0, 2.0, 3.0]);
    assert_float_relative_eq!(m.get_data_row_leading()[0][0], 1.0, 0.00001);
    assert_float_relative_eq!(m.get_data_row_leading()[2][0], 3.0, 0.00001);
}

#[test]
fn test_from_row_vector() {
    let m = Matrix::from_row_vector(&vec![1.0, 2.0, 3.0]);
    println!("{:?}", m.get_data_row_leading());
    assert_float_relative_eq!(m.get_data_row_leading()[0][0], 1.0, 0.00001);
    assert_float_relative_eq!(m.get_data_row_leading()[0][2], 3.0, 0.00001);
}

#[test]
fn test_get_column() {
    let m = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let col = m.get_column(1);
    assert_float_relative_eq!(col[0], 2.0, 0.00001);
    assert_float_relative_eq!(col[1], 5.0, 0.00001);
}

#[test]
fn test_get_row() {
    let m = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let row = m.get_row(1);
    assert_float_relative_eq!(row[0], 4.0, 0.00001);
    assert_float_relative_eq!(row[1], 5.0, 0.00001);
    assert_float_relative_eq!(row[2], 6.0, 0.00001);
}

// #[test]
// fn test_map() {
//     let m = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);

//     let mapped = m.map(|x| x * 2.0);

//     assert_float_relative_eq!(mapped.get_data_row_leading()[0][0], 2.0, 0.00001);
//     assert_float_relative_eq!(mapped.get_data_row_leading()[1][2], 12.0, 0.00001);
// }

#[test]
fn test_dot() {
    let m1 = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let m2 = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0], vec![3.0, 4.0], vec![5.0, 6.0]]);
    let result = m1.dot(&m2);
    assert_eq!(result.dim(), (2, 2));
    assert_float_relative_eq!(result.get_data_row_leading()[0][0], 22.0, 0.00001);
    assert_float_relative_eq!(result.get_data_row_leading()[1][1], 64.0, 0.00001);
}

#[test]
fn test_columns_sum() {
    let m = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let sums = m.columns_sum().get_data_row_leading();
    assert_float_relative_eq!(sums[0][0], 6.0, 0.00001);
    assert_float_relative_eq!(sums[1][0], 15.0, 0.00001);
}

#[test]
fn test_component_mul() {
    let m1 = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let m2 = Matrix::from_row_leading_vector2(&vec![vec![2.0, 3.0, 4.0], vec![5.0, 6.0, 7.0]]);
    let result = m1.component_mul(&m2);
    assert_float_relative_eq!(result.get_data_row_leading()[0][0], 2.0, 0.00001);
    assert_float_relative_eq!(result.get_data_row_leading()[1][2], 42.0, 0.00001);
}

#[test]
fn test_component_add() {
    let m1 = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let m2 = Matrix::from_row_leading_vector2(&vec![vec![2.0, 3.0, 4.0], vec![5.0, 6.0, 7.0]]);
    let result = m1.component_add(&m2);

    println!("{:#?}", result.get_data_row_leading());

    assert_float_relative_eq!(result.get_data_row_leading()[0][0], 3.0, 0.00001);
    assert_float_relative_eq!(result.get_data_row_leading()[1][2], 13.0, 0.00001);
}

#[test]
fn test_component_sub() {
    let m1 = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let m2 = Matrix::from_row_leading_vector2(&vec![vec![2.0, 3.0, 4.0], vec![5.0, 6.0, 7.0]]);
    let result = m1.component_sub(&m2);
    assert_float_relative_eq!(result.get_data_row_leading()[0][0], -1.0, 0.00001);
    assert_float_relative_eq!(result.get_data_row_leading()[1][2], -1.0, 0.00001);
}

#[test]
fn test_component_div() {
    let m1 = Matrix::from_row_leading_vector2(&vec![vec![2.0, 4.0, 6.0], vec![8.0, 10.0, 12.0]]);
    let m2 = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let result = m1.component_div(&m2);
    assert_float_relative_eq!(result.get_data_row_leading()[0][0], 2.0, 0.00001);
    assert_float_relative_eq!(result.get_data_row_leading()[1][2], 2.0, 0.00001);
}

#[test]
fn test_transpose() {
    let m = Matrix::from_row_leading_vector2(&vec![vec![1.0, 2.0, 3.0], vec![4.0, 5.0, 6.0]]);
    let transposed = m.transpose();
    assert_float_relative_eq!(transposed.get_data_row_leading()[0][0], 1.0, 0.00001);
    assert_float_relative_eq!(transposed.get_data_row_leading()[2][1], 6.0, 0.00001);
}