edgevec 0.9.0

High-performance embedded vector database for Browser, Node, and Edge
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

//! Scalar implementations for distance metrics.

/// Euclidean distance for f32 vectors (Scalar fallback).
///
/// Computes `sqrt(sum((a[i] - b[i])^2))`.
///
/// # Panics
///
/// Panics if `a` and `b` have different lengths.
///
/// # Example
///
/// ```
/// use edgevec::metric::scalar::euclidean_distance;
/// let a = vec![0.0f32, 0.0, 0.0];
/// let b = vec![3.0f32, 4.0, 0.0];
/// let dist = euclidean_distance(&a, &b);
/// assert!((dist - 5.0).abs() < 1e-6); // 3-4-5 triangle
/// ```
#[inline]
#[must_use]
pub fn euclidean_distance(a: &[f32], b: &[f32]) -> f32 {
    assert_eq!(a.len(), b.len());
    let mut sum: f32 = 0.0;
    for i in 0..a.len() {
        let diff = a[i] - b[i];
        sum += diff * diff;
    }
    sum.sqrt()
}

/// L2 Squared distance for f32 vectors (Scalar fallback).
///
/// Computes `sum((a[i] - b[i])^2)` without the square root.
///
/// # Panics
///
/// Panics if `a` and `b` have different lengths.
///
/// # Example
///
/// ```
/// use edgevec::metric::scalar::l2_squared;
/// let a = vec![1.0f32, 2.0, 3.0];
/// let b = vec![4.0f32, 2.0, 1.0];
/// // (1-4)^2 + (2-2)^2 + (3-1)^2 = 9 + 0 + 4 = 13
/// let dist_sq = l2_squared(&a, &b);
/// assert!((dist_sq - 13.0).abs() < 1e-6);
/// ```
#[inline]
#[must_use]
pub fn l2_squared(a: &[f32], b: &[f32]) -> f32 {
    assert_eq!(a.len(), b.len());
    let mut sum: f32 = 0.0;
    for i in 0..a.len() {
        let diff = a[i] - b[i];
        sum += diff * diff;
    }
    sum
}

/// L2 Squared distance for u8 vectors (Scalar fallback).
///
/// # Panics
///
/// Panics if `a` and `b` have different lengths.
#[inline]
#[must_use]
pub fn l2_squared_u8(a: &[u8], b: &[u8]) -> u32 {
    assert_eq!(a.len(), b.len());
    let mut sum: u32 = 0;
    for i in 0..a.len() {
        // Safe upcast to i32 to avoid overflow during subtraction
        let diff = i32::from(a[i]) - i32::from(b[i]);
        // SAFETY: diff*diff is always non-negative, so cast to u32 is safe
        #[allow(clippy::cast_sign_loss)]
        let sq = (diff * diff) as u32;
        sum += sq;
    }
    sum
}

/// Dot Product for u8 vectors (Scalar fallback).
///
/// # Panics
///
/// Panics if `a` and `b` have different lengths.
#[inline]
#[must_use]
pub fn dot_product_u8(a: &[u8], b: &[u8]) -> u32 {
    assert_eq!(a.len(), b.len());
    let mut sum: u32 = 0;
    for i in 0..a.len() {
        sum += u32::from(a[i]) * u32::from(b[i]);
    }
    sum
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_l2_squared_u8_scalar() {
        let a = vec![1, 2, 3];
        let b = vec![4, 2, 1];
        // (1-4)^2 + (2-2)^2 + (3-1)^2 = 9 + 0 + 4 = 13
        assert_eq!(l2_squared_u8(&a, &b), 13);
    }

    #[test]
    fn test_dot_product_u8_scalar() {
        let a = vec![1, 2, 3];
        let b = vec![4, 2, 1];
        // 1*4 + 2*2 + 3*1 = 4 + 4 + 3 = 11
        assert_eq!(dot_product_u8(&a, &b), 11);
    }

    #[test]
    fn test_overflow_protection() {
        // 255 * 255 = 65025.
        // 100 dimensions -> 6,502,500. Fits in u32 (4B).
        // 100k dimensions -> 6.5e9. Overflow u32 (4.29e9).
        // Max u32 is 4,294,967,295.
        // 65025 * N < 4,294,967,295 => N < 66050.
        // So for vectors longer than ~66k, we might overflow u32.
        // But HNSW usually uses 128-2048 dims. So safe.
        // Constraint in prompt: "Overflow: Result fits in u32 (max 50M < 4B)".
        // 50M refers to max possible value?
        // If dims=2048 (large), max val = 2048 * 255^2 = 133M. Fits comfortably.

        let n = 1000;
        let a = vec![255; n];
        let b = vec![0; n];
        // 255^2 * 1000 = 65,025,000
        assert_eq!(l2_squared_u8(&a, &b), 65_025_000);
    }

    #[test]
    fn test_euclidean_distance_scalar() {
        // 3-4-5 triangle
        let a = vec![0.0f32, 0.0, 0.0];
        let b = vec![3.0f32, 4.0, 0.0];
        let dist = euclidean_distance(&a, &b);
        assert!((dist - 5.0).abs() < 1e-6);
    }

    #[test]
    fn test_euclidean_distance_single_element() {
        let a = vec![5.0f32];
        let b = vec![3.0f32];
        // sqrt((5-3)^2) = sqrt(4) = 2
        let dist = euclidean_distance(&a, &b);
        assert!((dist - 2.0).abs() < 1e-6);
    }

    #[test]
    fn test_euclidean_distance_identical() {
        let a = vec![1.0f32, 2.0, 3.0, 4.0];
        let b = vec![1.0f32, 2.0, 3.0, 4.0];
        let dist = euclidean_distance(&a, &b);
        assert!(dist.abs() < 1e-6);
    }

    #[test]
    fn test_l2_squared_f32_scalar() {
        let a = vec![1.0f32, 2.0, 3.0];
        let b = vec![4.0f32, 2.0, 1.0];
        // (1-4)^2 + (2-2)^2 + (3-1)^2 = 9 + 0 + 4 = 13
        let dist = l2_squared(&a, &b);
        assert!((dist - 13.0).abs() < 1e-6);
    }
}