aprender-core 0.29.2

Next-generation machine learning library in pure Rust
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

pub(crate) use super::*;

#[test]
fn test_embedding_space_creation() {
    let space = EmbeddingSpace::new(10, 8);
    assert_eq!(space.num_entities(), 10);
    assert_eq!(space.dim(), 8);
}

#[test]
fn test_relation_matrix() {
    let mut space = EmbeddingSpace::new(5, 4);
    space.add_relation("knows");

    let matrix = space.get_relation_matrix("knows");
    assert!(matrix.is_some());
    assert_eq!(matrix.unwrap().len(), 4);
}

#[test]
fn test_bilinear_scoring() {
    let mut space = EmbeddingSpace::new(3, 4);
    space.add_relation("likes");

    let score = space.score(0, "likes", 1);
    assert!(score.is_finite());
}

#[test]
fn test_relation_composition() {
    let mut space = EmbeddingSpace::new(3, 4);
    space.add_relation("parent");

    let composed = space.compose_relations(&["parent", "parent"]);
    assert_eq!(composed.len(), 4);
    assert_eq!(composed[0].len(), 4);
}

#[test]
fn test_rescal_factorization() {
    let factorizer = RescalFactorizer::new(5, 4, 2);

    let triples = vec![(0, 0, 1), (1, 0, 2), (2, 1, 3)];

    let result = factorizer.factorize(&triples, 5);

    assert_eq!(result.entity_embeddings.len(), 5);
    assert_eq!(result.relation_cores.len(), 2);
}

#[test]
fn test_bilinear_scorer_predictions() {
    let mut space = EmbeddingSpace::new(5, 4);
    space.add_relation("knows");

    let scorer = BilinearScorer::new(space);
    let predictions = scorer.predict_tails(0, "knows", 3);

    assert_eq!(predictions.len(), 3);
}

// =========================================================================
// Additional coverage tests
// =========================================================================

#[test]
fn test_score_unknown_relation() {
    let space = EmbeddingSpace::new(3, 4);
    // No relations added
    let score = space.score(0, "unknown", 1);
    assert_eq!(score, 0.0);
}

#[test]
fn test_get_entity() {
    let space = EmbeddingSpace::new(5, 4);
    assert!(space.get_entity(0).is_some());
    assert!(space.get_entity(4).is_some());
    assert!(space.get_entity(5).is_none()); // Out of bounds
}

#[test]
fn test_set_entity() {
    let mut space = EmbeddingSpace::new(3, 4);
    let new_embedding = vec![1.0, 2.0, 3.0, 4.0];
    space.set_entity(0, new_embedding.clone());
    assert_eq!(space.get_entity(0).unwrap(), &new_embedding);
}

#[test]
fn test_set_entity_invalid_index() {
    let mut space = EmbeddingSpace::new(3, 4);
    let orig = space.get_entity(0).unwrap().clone();
    // Out of bounds - should do nothing
    space.set_entity(10, vec![1.0, 2.0, 3.0, 4.0]);
    // Entity 0 unchanged
    assert_eq!(space.get_entity(0).unwrap(), &orig);
}

#[test]
fn test_set_entity_wrong_dimension() {
    let mut space = EmbeddingSpace::new(3, 4);
    let orig = space.get_entity(0).unwrap().clone();
    // Wrong dimension (3 instead of 4) - should do nothing
    space.set_entity(0, vec![1.0, 2.0, 3.0]);
    assert_eq!(space.get_entity(0).unwrap(), &orig);
}

#[test]
fn test_compose_relations_empty() {
    let space = EmbeddingSpace::new(3, 4);
    let composed = space.compose_relations(&[]);
    // Should return zero matrix
    assert_eq!(composed.len(), 4);
    for row in &composed {
        for &val in row {
            assert_eq!(val, 0.0);
        }
    }
}

#[test]
fn test_compose_relations_unknown() {
    let space = EmbeddingSpace::new(3, 4);
    // Unknown relation - should return zero matrix
    let composed = space.compose_relations(&["unknown"]);
    assert_eq!(composed.len(), 4);
}

#[test]
fn test_compose_relations_mixed() {
    let mut space = EmbeddingSpace::new(3, 4);
    space.add_relation("parent");
    // One known, one unknown
    let composed = space.compose_relations(&["parent", "unknown"]);
    assert_eq!(composed.len(), 4);
}

#[test]
fn test_relation_matrix_wrapper() {
    let data = vec![vec![1.0, 2.0], vec![3.0, 4.0]];
    let rm = RelationMatrix::new(data.clone());
    assert_eq!(rm.len(), 2);
    assert!(!rm.is_empty());
    assert_eq!(rm.data, data);
}

#[test]
fn test_relation_matrix_empty() {
    let rm = RelationMatrix::new(vec![]);
    assert!(rm.is_empty());
    assert_eq!(rm.len(), 0);
}

#[test]
fn test_relation_matrix_debug_clone() {
    let rm = RelationMatrix::new(vec![vec![1.0]]);
    let debug_str = format!("{:?}", rm);
    assert!(debug_str.contains("RelationMatrix"));
    let cloned = rm.clone();
    assert_eq!(cloned.data, rm.data);
}

#[test]
fn test_bilinear_scorer_score_heads() {
    let mut space = EmbeddingSpace::new(5, 4);
    space.add_relation("knows");

    let scorer = BilinearScorer::new(space);
    let scores = scorer.score_heads("knows", 0);
    assert_eq!(scores.len(), 5);
}

#[test]
fn test_bilinear_scorer_score_tails() {
    let mut space = EmbeddingSpace::new(5, 4);
    space.add_relation("knows");

    let scorer = BilinearScorer::new(space);
    let scores = scorer.score_tails(0, "knows");
    assert_eq!(scores.len(), 5);
}

#[test]
fn test_bilinear_scorer_debug() {
    let space = EmbeddingSpace::new(3, 2);
    let scorer = BilinearScorer::new(space);
    let debug_str = format!("{:?}", scorer);
    assert!(debug_str.contains("BilinearScorer"));
}

#[test]
fn test_rescal_result_debug() {
    let factorizer = RescalFactorizer::new(3, 2, 1);
    let result = factorizer.factorize(&[(0, 0, 1)], 1);
    let debug_str = format!("{:?}", result);
    assert!(debug_str.contains("RescalResult"));
}

#[test]
fn test_rescal_factorizer_debug() {
    let factorizer = RescalFactorizer::new(3, 2, 1);
    let debug_str = format!("{:?}", factorizer);
    assert!(debug_str.contains("RescalFactorizer"));
}

#[test]
fn test_rescal_triples_out_of_bounds() {
    let factorizer = RescalFactorizer::new(3, 2, 1);
    // Triples with out-of-bounds indices should be ignored
    let result = factorizer.factorize(&[(0, 0, 1), (10, 0, 1), (0, 5, 1)], 2);
    assert_eq!(result.entity_embeddings.len(), 3);
}

#[test]
fn test_embedding_space_debug() {
    let space = EmbeddingSpace::new(2, 2);
    let debug_str = format!("{:?}", space);
    assert!(debug_str.contains("EmbeddingSpace"));
}

#[test]
fn test_matrix_multiply_edge_cases() {
    // Empty matrices
    let empty: Vec<Vec<f64>> = vec![];
    let result = matrix_multiply(&empty, &empty);
    assert!(result.is_empty());

    // Single element
    let a = vec![vec![2.0]];
    let b = vec![vec![3.0]];
    let result = matrix_multiply(&a, &b);
    assert!((result[0][0] - 6.0).abs() < 1e-10);
}