datarust 0.6.5

Scikit-learn-style preprocessing and classical ML 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
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
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238

//! Integration tests for `linear_model::LinearRegression` and `metrics`.

use datarust::linear_model::{LinearRegression, LinearSolver};
use datarust::metrics::regression::{
    explained_variance_score, max_error, mean_absolute_error, mean_squared_error, r2_score,
};
use datarust::traits::Predictor;
use datarust::Matrix;

fn approx(a: f64, b: f64, tol: f64) -> bool {
    (a - b).abs() < tol
}

/// Build a noise-free dataset y = Xβ + b for known β, b and recover them.
fn make_dataset(n: usize, beta: &[f64], intercept: f64) -> (Matrix, Vec<f64>) {
    // Deterministic, non-collinear features derived from the row index.
    let rows: Vec<Vec<f64>> = (0..n)
        .map(|i| {
            let i = i as f64;
            // p distinct basis functions — guarantees full column rank.
            let mut row = Vec::with_capacity(beta.len());
            for (j, &bj) in beta.iter().enumerate() {
                let _ = bj;
                let v = match j {
                    0 => i.sin(),
                    1 => (i + 7.0).ln(),
                    2 => (i * 0.3 + 1.0).exp() * 0.01,
                    _ => (i * (j as f64)).cos(),
                };
                row.push(v);
            }
            row
        })
        .collect();
    let y: Vec<f64> = rows
        .iter()
        .map(|r| {
            r.iter()
                .zip(beta.iter())
                .map(|(xj, &bj)| xj * bj)
                .sum::<f64>()
                + intercept
        })
        .collect();
    (Matrix::new(rows).unwrap(), y)
}

#[test]
fn recover_known_coefficients_cholesky() {
    let beta = vec![2.0, -3.5, 5.0];
    let (x, y) = make_dataset(40, &beta, 7.0);
    let mut model = LinearRegression::new();
    model.fit(&x, &y).unwrap();
    for (got, expected) in model.coef().iter().zip(beta.iter()) {
        assert!(
            approx(*got, *expected, 1e-6),
            "coef: got {got} expected {expected}"
        );
    }
    assert!(approx(model.intercept(), 7.0, 1e-6));
}

#[test]
fn recover_known_coefficients_svd() {
    let beta = vec![2.0, -3.5, 5.0];
    let (x, y) = make_dataset(40, &beta, 7.0);
    let mut model = LinearRegression::new().with_solver(LinearSolver::Svd);
    model.fit(&x, &y).unwrap();
    for (got, expected) in model.coef().iter().zip(beta.iter()) {
        assert!(
            approx(*got, *expected, 1e-6),
            "coef: got {got} expected {expected}"
        );
    }
    assert!(approx(model.intercept(), 7.0, 1e-6));
}

#[test]
fn solvers_agree_on_full_rank_data() {
    let beta = vec![1.5, -2.0, 0.3];
    let (x, y) = make_dataset(30, &beta, 4.0);
    let mut chol = LinearRegression::new();
    chol.fit(&x, &y).unwrap();
    let mut svd = LinearRegression::new().with_solver(LinearSolver::Svd);
    svd.fit(&x, &y).unwrap();
    for (c, s) in chol.coef().iter().zip(svd.coef().iter()) {
        assert!(
            approx(*c, *s, 1e-6),
            "solver disagreement: chol={c} svd={s}"
        );
    }
    assert!(approx(chol.intercept(), svd.intercept(), 1e-6));
}

#[test]
fn perfect_fit_yields_zero_residual() {
    let beta = vec![1.0, 2.0, -1.0];
    let (x, y) = make_dataset(25, &beta, 3.0);
    let mut model = LinearRegression::new();
    model.fit(&x, &y).unwrap();
    let pred = model.predict(&x).unwrap();
    let mse = mean_squared_error(&y, &pred, true).unwrap();
    assert!(mse < 1e-18, "mse too large: {mse}");
    let r2 = model.score(&x, &y).unwrap();
    assert!(approx(r2, 1.0, 1e-9));
}

#[test]
fn predict_on_held_out_data() {
    // y = 2x, fit_intercept=false.
    let train = Matrix::new(vec![vec![1.0], vec![2.0], vec![3.0], vec![4.0]]).unwrap();
    let y_train = vec![2.0, 4.0, 6.0, 8.0];
    let mut model = LinearRegression::new().with_fit_intercept(false);
    model.fit(&train, &y_train).unwrap();
    let held_out = Matrix::new(vec![vec![100.0]]).unwrap();
    let pred = model.predict(&held_out).unwrap();
    assert!(approx(pred[0], 200.0, 1e-6));
}

#[test]
fn score_returns_r2() {
    let beta = vec![1.0, 0.5];
    let (x, y) = make_dataset(30, &beta, 2.0);
    let mut model = LinearRegression::new();
    model.fit(&x, &y).unwrap();
    let score = model.score(&x, &y).unwrap();
    let pred = model.predict(&x).unwrap();
    let direct = r2_score(&y, &pred).unwrap();
    assert!(approx(score, direct, 1e-12));
}

#[test]
fn metrics_sklearn_parity() {
    let y_true = vec![3.0, -0.5, 2.0, 7.0];
    let y_pred = vec![2.5, 0.0, 2.0, 8.0];
    assert!(approx(
        mean_squared_error(&y_true, &y_pred, true).unwrap(),
        0.375,
        1e-12
    ));
    assert!(approx(
        mean_squared_error(&y_true, &y_pred, false).unwrap(),
        0.375_f64.sqrt(),
        1e-12
    ));
    assert!(approx(
        mean_absolute_error(&y_true, &y_pred).unwrap(),
        0.5,
        1e-12
    ));
    assert!(approx(
        r2_score(&y_true, &y_pred).unwrap(),
        0.9486081370449679,
        1e-9
    ));
    assert!(approx(max_error(&y_true, &y_pred).unwrap(), 1.0, 1e-12));
}

#[test]
fn metrics_perfect_predictions() {
    let y = vec![1.0, 2.0, 3.0, 4.0, 5.0];
    assert!(approx(r2_score(&y, &y).unwrap(), 1.0, 1e-12));
    assert!(approx(
        explained_variance_score(&y, &y).unwrap(),
        1.0,
        1e-12
    ));
    assert!(approx(
        mean_squared_error(&y, &y, true).unwrap(),
        0.0,
        1e-12
    ));
    assert!(approx(max_error(&y, &y).unwrap(), 0.0, 1e-12));
}

#[test]
fn constant_target_recovers_intercept() {
    let x = Matrix::new(vec![vec![1.0], vec![2.0], vec![5.0], vec![9.0]]).unwrap();
    let y = vec![7.0, 7.0, 7.0, 7.0];
    let mut model = LinearRegression::new();
    model.fit(&x, &y).unwrap();
    assert!(approx(model.intercept(), 7.0, 1e-9));
    assert!(model.coef()[0].abs() < 1e-9);
}

#[test]
fn predict_before_fit_errors() {
    let model = LinearRegression::new();
    let x = Matrix::new(vec![vec![1.0]]).unwrap();
    let err = model.predict(&x).unwrap_err();
    assert!(matches!(err, datarust::DatarustError::NotFitted(_)));
}

#[test]
fn predict_shape_mismatch_errors() {
    // Use non-collinear features so fit succeeds, then predict with wrong width.
    let rows: Vec<Vec<f64>> = (0..10)
        .map(|i| {
            let i = i as f64;
            vec![i.sin(), (i + 1.0).ln()]
        })
        .collect();
    let x = Matrix::new(rows).unwrap();
    let y: Vec<f64> = (0..10).map(|i| i as f64).collect();
    let mut model = LinearRegression::new();
    model.fit(&x, &y).unwrap();
    let bad = Matrix::new(vec![vec![1.0, 2.0, 3.0]]).unwrap();
    let err = model.predict(&bad).unwrap_err();
    assert!(matches!(err, datarust::DatarustError::ShapeMismatch { .. }));
}

#[test]
fn fit_y_length_mismatch_errors() {
    let x = Matrix::new(vec![vec![1.0], vec![2.0]]).unwrap();
    let mut model = LinearRegression::new();
    let err = model.fit(&x, &[1.0]).unwrap_err();
    assert!(matches!(err, datarust::DatarustError::ShapeMismatch { .. }));
}

#[test]
fn is_fitted_flag() {
    let x = Matrix::new(vec![vec![1.0], vec![2.0]]).unwrap();
    let mut model = LinearRegression::new();
    assert!(!model.is_fitted());
    model.fit(&x, &[2.0, 4.0]).unwrap();
    assert!(model.is_fitted());
}

#[test]
fn fit_predict_convenience() {
    let x = Matrix::new(vec![vec![1.0], vec![2.0], vec![3.0]]).unwrap();
    let y = vec![2.0, 4.0, 6.0];
    let mut model = LinearRegression::new().with_fit_intercept(false);
    let pred = model.fit_predict(&x, &y).unwrap();
    for (p, &t) in pred.iter().zip(y.iter()) {
        assert!(approx(*p, t, 1e-9));
    }
}