pmat 3.17.0

PMAT - Zero-config AI context generation and code quality toolkit (CLI, MCP, HTTP)
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
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339

#![cfg_attr(coverage_nightly, coverage(off))]
#[cfg(test)]
mod tests {
    use super::super::super::TursoVectorDB;
    use super::super::engine::ClusteringEngine;
    #[allow(unused_imports)]
    use super::super::types::{
        Cluster, ClusterFilters, ClusterMember, ClusterResult, ClusteringMethod, Dendrogram,
        DendrogramMerge, Linkage, OutlierPoint,
    };
    #[allow(unused_imports)]
    use std::collections::HashMap;
    use std::sync::Arc;

    /// Helper to create a clustering engine for tests
    async fn create_test_engine() -> ClusteringEngine {
        let db = TursoVectorDB::new_local(":memory:")
            .await
            .expect("Failed to create test database");
        ClusteringEngine::new(Arc::new(db))
    }

    // ==================== Hierarchical Clustering Tests ====================

    #[tokio::test]
    async fn test_hierarchical_basic() {
        let engine = create_test_engine().await;

        let vectors = vec![
            vec![0.0, 0.0],
            vec![1.0, 0.0],
            vec![10.0, 0.0],
            vec![11.0, 0.0],
        ];

        let dendrogram = engine.hierarchical(&vectors, Linkage::Single).unwrap();

        // Should have n-1 merges
        assert_eq!(dendrogram.merges.len(), 3);
    }

    #[tokio::test]
    async fn test_hierarchical_single_linkage() {
        let engine = create_test_engine().await;

        let vectors = vec![vec![0.0, 0.0], vec![1.0, 0.0], vec![2.0, 0.0]];

        let dendrogram = engine.hierarchical(&vectors, Linkage::Single).unwrap();

        assert_eq!(dendrogram.merges.len(), 2);
        // First merge should have distance ~1.0
        assert!((dendrogram.merges[0].distance - 1.0).abs() < 1e-6);
    }

    #[tokio::test]
    async fn test_hierarchical_complete_linkage() {
        let engine = create_test_engine().await;

        let vectors = vec![vec![0.0, 0.0], vec![1.0, 0.0], vec![2.0, 0.0]];

        let dendrogram = engine.hierarchical(&vectors, Linkage::Complete).unwrap();

        assert_eq!(dendrogram.merges.len(), 2);
    }

    #[tokio::test]
    async fn test_hierarchical_average_linkage() {
        let engine = create_test_engine().await;

        let vectors = vec![vec![0.0, 0.0], vec![1.0, 0.0], vec![2.0, 0.0]];

        let dendrogram = engine.hierarchical(&vectors, Linkage::Average).unwrap();

        assert_eq!(dendrogram.merges.len(), 2);
    }

    #[tokio::test]
    async fn test_hierarchical_empty_vectors() {
        let engine = create_test_engine().await;

        let vectors: Vec<Vec<f32>> = vec![];
        let result = engine.hierarchical(&vectors, Linkage::Single);

        assert!(result.is_err());
        assert_eq!(result.unwrap_err(), "Cannot cluster empty vector set");
    }

    #[tokio::test]
    async fn test_hierarchical_single_point() {
        let engine = create_test_engine().await;

        let vectors = vec![vec![1.0, 2.0, 3.0]];
        let dendrogram = engine.hierarchical(&vectors, Linkage::Single).unwrap();

        // No merges for single point
        assert_eq!(dendrogram.merges.len(), 0);
    }

    #[tokio::test]
    async fn test_hierarchical_two_points() {
        let engine = create_test_engine().await;

        let vectors = vec![vec![0.0, 0.0], vec![3.0, 4.0]];
        let dendrogram = engine.hierarchical(&vectors, Linkage::Single).unwrap();

        assert_eq!(dendrogram.merges.len(), 1);
        // Distance should be 5 (3-4-5 triangle)
        assert!((dendrogram.merges[0].distance - 5.0).abs() < 1e-6);
    }

    #[tokio::test]
    async fn test_hierarchical_merge_ordering() {
        let engine = create_test_engine().await;

        // Clusters: (0, 0), (0.1, 0) are very close
        // (10, 0) is far
        let vectors = vec![vec![0.0, 0.0], vec![0.1, 0.0], vec![10.0, 0.0]];

        let dendrogram = engine.hierarchical(&vectors, Linkage::Single).unwrap();

        assert_eq!(dendrogram.merges.len(), 2);
        // First merge should have small distance
        assert!(dendrogram.merges[0].distance < 1.0);
        // Second merge should have larger distance
        assert!(dendrogram.merges[1].distance > 1.0);
    }

    // ==================== DBSCAN Tests ====================

    #[tokio::test]
    async fn test_dbscan_basic() {
        let engine = create_test_engine().await;

        // Two clusters of 3 points each
        let vectors = vec![
            vec![0.0, 0.0],
            vec![0.1, 0.0],
            vec![0.0, 0.1],
            vec![10.0, 10.0],
            vec![10.1, 10.0],
            vec![10.0, 10.1],
        ];

        let labels = engine.dbscan(&vectors, 0.5, 2).unwrap();

        assert_eq!(labels.len(), 6);

        // First three points should be in same cluster
        assert_eq!(labels[0], labels[1]);
        assert_eq!(labels[1], labels[2]);
        assert!(labels[0] >= 0);

        // Last three points should be in same cluster
        assert_eq!(labels[3], labels[4]);
        assert_eq!(labels[4], labels[5]);
        assert!(labels[3] >= 0);

        // Different clusters
        assert_ne!(labels[0], labels[3]);
    }

    #[tokio::test]
    async fn test_dbscan_noise_points() {
        let engine = create_test_engine().await;

        // One cluster and one isolated noise point
        let vectors = vec![
            vec![0.0, 0.0],
            vec![0.1, 0.0],
            vec![0.0, 0.1],
            vec![100.0, 100.0], // Isolated point
        ];

        let labels = engine.dbscan(&vectors, 0.5, 2).unwrap();

        // Last point should be noise (-1)
        assert_eq!(labels[3], -1);

        // First three should be in a cluster
        assert_eq!(labels[0], labels[1]);
        assert_eq!(labels[1], labels[2]);
        assert!(labels[0] >= 0);
    }

    #[tokio::test]
    async fn test_dbscan_all_noise() {
        let engine = create_test_engine().await;

        // All points too far apart
        let vectors = vec![
            vec![0.0, 0.0],
            vec![10.0, 0.0],
            vec![20.0, 0.0],
            vec![30.0, 0.0],
        ];

        let labels = engine.dbscan(&vectors, 0.5, 2).unwrap();

        // All points should be noise
        assert!(labels.iter().all(|&l| l == -1));
    }

    #[tokio::test]
    async fn test_dbscan_single_cluster() {
        let engine = create_test_engine().await;

        // All points close together
        let vectors = vec![
            vec![0.0, 0.0],
            vec![0.1, 0.0],
            vec![0.2, 0.0],
            vec![0.3, 0.0],
        ];

        let labels = engine.dbscan(&vectors, 0.5, 2).unwrap();

        // All points should be in same cluster
        assert!(labels.iter().all(|&l| l == labels[0]));
        assert!(labels[0] >= 0);
    }

    #[tokio::test]
    async fn test_dbscan_empty_vectors() {
        let engine = create_test_engine().await;

        let vectors: Vec<Vec<f32>> = vec![];
        let result = engine.dbscan(&vectors, 0.5, 2);

        assert!(result.is_err());
        assert_eq!(result.unwrap_err(), "Cannot cluster empty vector set");
    }

    #[tokio::test]
    async fn test_dbscan_large_epsilon() {
        let engine = create_test_engine().await;

        // With large epsilon, all points should be in one cluster
        let vectors = vec![vec![0.0, 0.0], vec![5.0, 0.0], vec![10.0, 0.0]];

        let labels = engine.dbscan(&vectors, 100.0, 2).unwrap();

        // All in same cluster
        assert_eq!(labels[0], labels[1]);
        assert_eq!(labels[1], labels[2]);
        assert!(labels[0] >= 0);
    }

    #[tokio::test]
    async fn test_dbscan_high_min_samples() {
        let engine = create_test_engine().await;

        // With high min_samples, all become noise
        let vectors = vec![vec![0.0, 0.0], vec![0.1, 0.0], vec![0.2, 0.0]];

        let labels = engine.dbscan(&vectors, 0.5, 10).unwrap();

        // All noise since min_samples > number of points
        assert!(labels.iter().all(|&l| l == -1));
    }

    // ==================== Silhouette Score Tests ====================

    #[tokio::test]
    async fn test_silhouette_score_perfect_clusters() {
        let engine = create_test_engine().await;

        // Two perfectly separated clusters
        let vectors = vec![
            vec![0.0, 0.0],
            vec![0.01, 0.01],
            vec![100.0, 100.0],
            vec![100.01, 100.01],
        ];
        let labels = vec![0, 0, 1, 1];

        let score = engine.compute_silhouette_score(&vectors, &labels);

        // Should be close to 1.0 for well-separated clusters
        assert!(score > 0.9);
    }

    #[tokio::test]
    async fn test_silhouette_score_overlapping_clusters() {
        let engine = create_test_engine().await;

        // Overlapping clusters - poor separation
        let vectors = vec![
            vec![0.0, 0.0],
            vec![1.0, 1.0],
            vec![0.5, 0.5],
            vec![1.5, 1.5],
        ];
        let labels = vec![0, 0, 1, 1];

        let score = engine.compute_silhouette_score(&vectors, &labels);

        // Score should be lower due to overlap
        assert!(score < 0.9);
    }

    #[tokio::test]
    async fn test_silhouette_score_empty_vectors() {
        let engine = create_test_engine().await;

        let vectors: Vec<Vec<f32>> = vec![];
        let labels: Vec<usize> = vec![];

        let score = engine.compute_silhouette_score(&vectors, &labels);

        assert_eq!(score, 0.0);
    }

    #[tokio::test]
    async fn test_silhouette_score_single_point() {
        let engine = create_test_engine().await;

        let vectors = vec![vec![1.0, 2.0, 3.0]];
        let labels = vec![0];

        let score = engine.compute_silhouette_score(&vectors, &labels);

        // Single point: a=0, b=MAX, silhouette = 1.0 - (0/MAX) = 1.0
        assert!(score > 0.9);
    }

    #[tokio::test]
    async fn test_silhouette_score_single_cluster() {
        let engine = create_test_engine().await;

        let vectors = vec![vec![0.0, 0.0], vec![1.0, 0.0], vec![2.0, 0.0]];
        let labels = vec![0, 0, 0];

        let score = engine.compute_silhouette_score(&vectors, &labels);

        // Single cluster: nearest_cluster_distance returns MAX
        // silhouette = 1.0 - (a / MAX) ≈ 1.0
        assert!(score > 0.9);
    }
}