plotters-statistical 0.1.0

Statistical chart primitives (box, violin, ROC, PR, regularization-path, residual) as native plotters series
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

//! Unit tests for the `stats` layer against hand-computed reference values and
//! every edge case called out in Milestone 1.

use approx::assert_abs_diff_eq;
use plotters_statistical::stats::{
    kde_curve, precision_recall_curve, quartiles, roc::auc_trapezoid, roc_curve,
    silverman_bandwidth, StatsError,
};

// ---------------------------------------------------------------------------
// Quartiles (type-7 / linear interpolation)
// ---------------------------------------------------------------------------

#[test]
fn quartiles_simple_odd() {
    let q = quartiles(&[1.0, 2.0, 3.0, 4.0, 5.0]).unwrap();
    assert_abs_diff_eq!(q.q1, 2.0);
    assert_abs_diff_eq!(q.median, 3.0);
    assert_abs_diff_eq!(q.q3, 4.0);
    assert_abs_diff_eq!(q.iqr, 2.0);
    assert_abs_diff_eq!(q.lower_whisker, 1.0);
    assert_abs_diff_eq!(q.upper_whisker, 5.0);
    assert!(q.outliers.is_empty());
}

#[test]
fn quartiles_with_high_outlier() {
    // Hand-computed for [1,2,3,4,5,100] under type-7:
    //   q1 = 2.25, median = 3.5, q3 = 4.75, IQR = 2.5
    //   upper fence = 4.75 + 1.5*2.5 = 8.5  -> 100 is an outlier
    //   upper whisker = 5 (largest value <= 8.5)
    let q = quartiles(&[1.0, 2.0, 3.0, 4.0, 5.0, 100.0]).unwrap();
    assert_abs_diff_eq!(q.q1, 2.25);
    assert_abs_diff_eq!(q.median, 3.5);
    assert_abs_diff_eq!(q.q3, 4.75);
    assert_abs_diff_eq!(q.iqr, 2.5);
    assert_abs_diff_eq!(q.upper_fence(), 8.5);
    assert_abs_diff_eq!(q.upper_whisker, 5.0);
    assert_abs_diff_eq!(q.lower_whisker, 1.0);
    assert_eq!(q.outliers, vec![100.0]);
}

#[test]
fn quartiles_single_point() {
    let q = quartiles(&[7.0]).unwrap();
    assert_abs_diff_eq!(q.q1, 7.0);
    assert_abs_diff_eq!(q.median, 7.0);
    assert_abs_diff_eq!(q.q3, 7.0);
    assert_abs_diff_eq!(q.iqr, 0.0);
    assert!(q.outliers.is_empty());
}

#[test]
fn quartiles_all_identical_zero_iqr() {
    let q = quartiles(&[3.0, 3.0, 3.0, 3.0]).unwrap();
    assert_abs_diff_eq!(q.iqr, 0.0);
    // Zero IQR collapses the fences onto the value: nothing is an outlier.
    assert!(q.outliers.is_empty());
    assert_abs_diff_eq!(q.lower_whisker, 3.0);
    assert_abs_diff_eq!(q.upper_whisker, 3.0);
}

#[test]
fn quartiles_empty_errors() {
    assert_eq!(quartiles(&[]), Err(StatsError::EmptyInput));
}

// ---------------------------------------------------------------------------
// KDE
// ---------------------------------------------------------------------------

#[test]
fn silverman_positive_for_spread_data() {
    let h = silverman_bandwidth(&[1.0, 2.0, 3.0, 4.0, 5.0, 6.0]).unwrap();
    assert!(h > 0.0 && h.is_finite());
}

#[test]
fn silverman_errors_on_degenerate() {
    assert_eq!(
        silverman_bandwidth(&[5.0]),
        Err(StatsError::InvalidBandwidth)
    );
    assert_eq!(
        silverman_bandwidth(&[2.0, 2.0, 2.0]),
        Err(StatsError::InvalidBandwidth)
    );
}

#[test]
fn kde_integrates_to_one() {
    // A symmetric sample: the density should integrate (trapezoid) to ~1 and
    // peak near the center of mass.
    let data = [-2.0, -1.0, 0.0, 1.0, 2.0];
    let curve = kde_curve(&data, Some(0.5), 400, 4.0).unwrap();
    let area: f64 = curve
        .xs
        .windows(2)
        .zip(curve.density.windows(2))
        .map(|(x, d)| (x[1] - x[0]) * (d[0] + d[1]) / 2.0)
        .sum();
    assert_abs_diff_eq!(area, 1.0, epsilon = 1e-2);
    assert!(curve.density.iter().all(|&d| d >= 0.0));

    // Peak index should sit near x = 0 for this symmetric data.
    let (peak_i, _) = curve
        .density
        .iter()
        .enumerate()
        .max_by(|a, b| a.1.partial_cmp(b.1).unwrap())
        .unwrap();
    assert_abs_diff_eq!(curve.xs[peak_i], 0.0, epsilon = 0.1);
}

#[test]
fn kde_rejects_bad_bandwidth() {
    assert_eq!(
        kde_curve(&[1.0, 2.0, 3.0], Some(0.0), 100, 3.0),
        Err(StatsError::InvalidBandwidth)
    );
    assert_eq!(kde_curve(&[], None, 100, 3.0), Err(StatsError::EmptyInput));
}

// ---------------------------------------------------------------------------
// ROC
// ---------------------------------------------------------------------------

#[test]
fn roc_matches_reference() {
    // The scikit-learn documentation example:
    //   y_true  = [0, 0, 1, 1], y_score = [0.1, 0.4, 0.35, 0.8]
    //   fpr = [0, 0, 0.5, 0.5, 1], tpr = [0, 0.5, 0.5, 1, 1], AUC = 0.75
    let scores = [0.1, 0.4, 0.35, 0.8];
    let labels = [false, false, true, true];
    let roc = roc_curve(&scores, &labels).unwrap();
    assert_abs_diff_eq!(roc.auc, 0.75);

    let fprs: Vec<f64> = roc.points.iter().map(|p| p.fpr).collect();
    let tprs: Vec<f64> = roc.points.iter().map(|p| p.tpr).collect();
    assert_eq!(fprs, vec![0.0, 0.0, 0.5, 0.5, 1.0]);
    assert_eq!(tprs, vec![0.0, 0.5, 0.5, 1.0, 1.0]);
}

#[test]
fn roc_perfect_separation_auc_one() {
    let scores = [0.9, 0.8, 0.2, 0.1];
    let labels = [true, true, false, false];
    let roc = roc_curve(&scores, &labels).unwrap();
    assert_abs_diff_eq!(roc.auc, 1.0);
}

#[test]
fn roc_edge_cases_error() {
    assert_eq!(
        roc_curve(&[0.1, 0.2], &[true]),
        Err(StatsError::LengthMismatch {
            scores: 2,
            labels: 1
        })
    );
    assert_eq!(roc_curve(&[], &[]), Err(StatsError::EmptyInput));
    assert_eq!(
        roc_curve(&[0.1, 0.2], &[true, true]),
        Err(StatsError::NoNegativeLabels)
    );
    assert_eq!(
        roc_curve(&[0.1, 0.2], &[false, false]),
        Err(StatsError::NoPositiveLabels)
    );
}

#[test]
fn auc_trapezoid_unit_triangle() {
    // Straight line from (0,0) to (1,1) -> area 0.5.
    assert_abs_diff_eq!(auc_trapezoid(&[(0.0, 0.0), (1.0, 1.0)]), 0.5);
}

// ---------------------------------------------------------------------------
// Precision / recall
// ---------------------------------------------------------------------------

#[test]
fn pr_matches_reference() {
    // Same data; scikit-learn average_precision_score = 0.8333...
    let scores = [0.1, 0.4, 0.35, 0.8];
    let labels = [false, false, true, true];
    let pr = precision_recall_curve(&scores, &labels).unwrap();
    assert_abs_diff_eq!(pr.average_precision, 5.0 / 6.0, epsilon = 1e-9);
    assert_abs_diff_eq!(pr.baseline, 0.5);
    // Recall is non-decreasing and reaches 1.0.
    assert!(pr.points.windows(2).all(|w| w[1].recall >= w[0].recall));
    assert_abs_diff_eq!(pr.points.last().unwrap().recall, 1.0);
}

#[test]
fn pr_no_positives_errors() {
    assert_eq!(
        precision_recall_curve(&[0.1, 0.2], &[false, false]),
        Err(StatsError::NoPositiveLabels)
    );
}