yscv-tensor 0.1.8

SIMD-accelerated tensor library with f32/f16/bf16 support and 80+ operations
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

use crate::Tensor;

#[test]
fn sum_reduces_all_elements() {
    let tensor = Tensor::from_vec(vec![4], vec![1.0, 2.0, 3.0, 4.0]).unwrap();
    assert_eq!(tensor.sum(), 10.0);
}

#[test]
fn sum_axis_reduces_requested_axis() {
    let tensor = Tensor::from_vec(vec![2, 3], vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0]).unwrap();

    let axis0 = tensor.sum_axis(0).unwrap();
    assert_eq!(axis0.shape(), &[3]);
    assert_eq!(axis0.data(), &[5.0, 7.0, 9.0]);

    let axis1 = tensor.sum_axis(1).unwrap();
    assert_eq!(axis1.shape(), &[2]);
    assert_eq!(axis1.data(), &[6.0, 15.0]);
}

#[test]
fn sum_axis_rejects_invalid_axis() {
    use crate::TensorError;
    let tensor = Tensor::zeros(vec![2, 3]).unwrap();
    let err = tensor.sum_axis(2).unwrap_err();
    assert_eq!(err, TensorError::InvalidAxis { axis: 2, rank: 2 });
}

#[test]
fn mean_reduces_all_elements() {
    let t = Tensor::from_vec(vec![4], vec![1.0, 2.0, 3.0, 4.0]).unwrap();
    assert_eq!(t.mean(), 2.5);
}

#[test]
fn mean_empty_returns_nan() {
    let t = Tensor::from_vec(vec![0], vec![]).unwrap();
    assert!(t.mean().is_nan());
}

#[test]
fn max_value_and_min_value() {
    let t = Tensor::from_vec(vec![4], vec![3.0, 1.0, 4.0, 2.0]).unwrap();
    assert_eq!(t.max_value(), 4.0);
    assert_eq!(t.min_value(), 1.0);
}

#[test]
fn argmax_and_argmin() {
    let t = Tensor::from_vec(vec![4], vec![3.0, 1.0, 4.0, 2.0]).unwrap();
    assert_eq!(t.argmax(), Some(2));
    assert_eq!(t.argmin(), Some(1));
}

#[test]
fn argmax_empty_returns_none() {
    let t = Tensor::from_vec(vec![0], vec![]).unwrap();
    assert_eq!(t.argmax(), None);
    assert_eq!(t.argmin(), None);
}

#[test]
fn var_and_std_dev() {
    let t = Tensor::from_vec(vec![4], vec![2.0, 4.0, 4.0, 4.0]).unwrap();
    assert!((t.mean() - 3.5).abs() < 1e-5);
    assert!((t.var() - 0.75).abs() < 1e-5);
    assert!((t.std_dev() - 0.75_f32.sqrt()).abs() < 1e-5);
}

#[test]
fn mean_axis_reduces_requested_axis() {
    let t = Tensor::from_vec(vec![2, 3], vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0]).unwrap();
    let m0 = t.mean_axis(0).unwrap();
    assert_eq!(m0.shape(), &[3]);
    assert_eq!(m0.data(), &[2.5, 3.5, 4.5]);

    let m1 = t.mean_axis(1).unwrap();
    assert_eq!(m1.shape(), &[2]);
    assert_eq!(m1.data(), &[2.0, 5.0]);
}

#[test]
fn max_axis_and_min_axis() {
    let t = Tensor::from_vec(vec![2, 3], vec![1.0, 5.0, 3.0, 4.0, 2.0, 6.0]).unwrap();
    let mx0 = t.max_axis(0).unwrap();
    assert_eq!(mx0.data(), &[4.0, 5.0, 6.0]);
    let mn1 = t.min_axis(1).unwrap();
    assert_eq!(mn1.data(), &[1.0, 2.0]);
}

#[test]
fn var_axis_computes_population_variance() {
    let t = Tensor::from_vec(vec![2, 2], vec![1.0, 3.0, 5.0, 7.0]).unwrap();
    let v = t.var_axis(0).unwrap();
    assert_eq!(v.shape(), &[2]);
    assert!((v.data()[0] - 4.0).abs() < 1e-5);
    assert!((v.data()[1] - 4.0).abs() < 1e-5);
}

#[test]
fn median_odd_count() {
    let t = Tensor::from_vec(vec![5], vec![3.0, 1.0, 4.0, 1.0, 5.0]).unwrap();
    assert_eq!(t.median(), 3.0);
}

#[test]
fn median_even_count() {
    let t = Tensor::from_vec(vec![4], vec![1.0, 2.0, 3.0, 4.0]).unwrap();
    assert_eq!(t.median(), 2.5);
}

#[test]
fn median_axis_removes_dim() {
    let t = Tensor::from_vec(vec![2, 3], vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0]).unwrap();
    let m = t.median_axis(1).unwrap();
    assert_eq!(m.shape(), &[2]);
    assert_eq!(m.data()[0], 2.0); // median of [1,2,3]
    assert_eq!(m.data()[1], 5.0); // median of [4,5,6]
}

#[test]
fn any_with_zeros_and_nonzeros() {
    let t1 = Tensor::from_vec(vec![3], vec![0.0, 0.0, 0.0]).unwrap();
    assert!(!t1.any());
    let t2 = Tensor::from_vec(vec![3], vec![0.0, 1.0, 0.0]).unwrap();
    assert!(t2.any());
}

#[test]
fn all_with_zeros_and_nonzeros() {
    let t1 = Tensor::from_vec(vec![3], vec![1.0, 2.0, 3.0]).unwrap();
    assert!(t1.all());
    let t2 = Tensor::from_vec(vec![3], vec![1.0, 0.0, 3.0]).unwrap();
    assert!(!t2.all());
}

#[test]
fn quantile_matches_median() {
    let t = Tensor::from_vec(vec![5], vec![1.0, 2.0, 3.0, 4.0, 5.0]).unwrap();
    assert!((t.quantile(0.5) - t.median()).abs() < 1e-6);
}

#[test]
fn quantile_extremes() {
    let t = Tensor::from_vec(vec![5], vec![10.0, 20.0, 30.0, 40.0, 50.0]).unwrap();
    assert!((t.quantile(0.0) - 10.0).abs() < 1e-6);
    assert!((t.quantile(1.0) - 50.0).abs() < 1e-6);
}

#[test]
fn quantile_interpolation() {
    let t = Tensor::from_vec(vec![4], vec![1.0, 2.0, 3.0, 4.0]).unwrap();
    let q25 = t.quantile(0.25);
    assert!((q25 - 1.75).abs() < 1e-5);
}