evo-rs 0.1.0

Evolutionary computation in Rust — genetic algorithms, differential evolution, NSGA-II multi-objective, genetic programming
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
202
203

use rand::Rng;
use crate::individual::Individual;

/// Uniform crossover: each gene is taken from either parent with equal probability.
pub fn uniform_crossover(p1: &Individual, p2: &Individual, rng: &mut impl Rng) -> (Individual, Individual) {
    let genome1: Vec<f64> = p1.genome.iter().zip(p2.genome.iter())
        .map(|(&a, &b)| if rng.gen::<bool>() { a } else { b })
        .collect();
    let genome2: Vec<f64> = p1.genome.iter().zip(p2.genome.iter())
        .map(|(&a, &b)| if rng.gen::<bool>() { a } else { b })
        .collect();
    (Individual::new(genome1), Individual::new(genome2))
}

/// Single-point crossover.
pub fn single_point_crossover(p1: &Individual, p2: &Individual, rng: &mut impl Rng) -> (Individual, Individual) {
    let len = p1.genome.len().min(p2.genome.len());
    if len == 0 {
        return (p1.clone(), p2.clone());
    }
    let point = rng.gen_range(1..len);
    let mut g1 = p1.genome.clone();
    let mut g2 = p2.genome.clone();
    for i in point..len {
        g1[i] = p2.genome[i];
        g2[i] = p1.genome[i];
    }
    (Individual::new(g1), Individual::new(g2))
}

/// Two-point crossover.
pub fn two_point_crossover(p1: &Individual, p2: &Individual, rng: &mut impl Rng) -> (Individual, Individual) {
    let len = p1.genome.len().min(p2.genome.len());
    if len < 2 {
        return single_point_crossover(p1, p2, rng);
    }
    let mut pt1 = rng.gen_range(1..len);
    let mut pt2 = rng.gen_range(1..len);
    if pt1 > pt2 {
        std::mem::swap(&mut pt1, &mut pt2);
    }
    let mut g1 = p1.genome.clone();
    let mut g2 = p2.genome.clone();
    for i in pt1..=pt2.min(len - 1) {
        g1[i] = p2.genome[i];
        g2[i] = p1.genome[i];
    }
    (Individual::new(g1), Individual::new(g2))
}

/// Blend crossover (BLX-alpha). For each gene, the child value is sampled from
/// an extended interval [min - alpha*range, max + alpha*range].
pub fn blend_crossover(p1: &Individual, p2: &Individual, alpha: f64, rng: &mut impl Rng) -> (Individual, Individual) {
    let genome1: Vec<f64> = p1.genome.iter().zip(p2.genome.iter())
        .map(|(&a, &b)| {
            let lo = a.min(b);
            let hi = a.max(b);
            let range = hi - lo;
            rng.gen_range((lo - alpha * range)..=(hi + alpha * range))
        })
        .collect();
    let genome2: Vec<f64> = p1.genome.iter().zip(p2.genome.iter())
        .map(|(&a, &b)| {
            let lo = a.min(b);
            let hi = a.max(b);
            let range = hi - lo;
            rng.gen_range((lo - alpha * range)..=(hi + alpha * range))
        })
        .collect();
    (Individual::new(genome1), Individual::new(genome2))
}

/// Simulated binary crossover (SBX). Commonly used in NSGA-II.
pub fn simulated_binary_crossover(
    p1: &Individual,
    p2: &Individual,
    eta: f64,
    bounds: &[(f64, f64)],
    rng: &mut impl Rng,
) -> (Individual, Individual) {
    let n = p1.genome.len();
    let mut g1 = vec![0.0; n];
    let mut g2 = vec![0.0; n];
    for j in 0..n {
        let a = p1.genome[j];
        let b = p2.genome[j];
        if (a - b).abs() < 1e-14 {
            g1[j] = a;
            g2[j] = b;
            continue;
        }
        let (y1, y2) = if a < b { (a, b) } else { (b, a) };
        let (lo, hi) = bounds.get(j).copied().unwrap_or((f64::NEG_INFINITY, f64::INFINITY));
        let rand_val = rng.gen::<f64>();
        let beta_q = if rand_val <= 0.5 {
            (2.0 * rand_val).powf(1.0 / (eta + 1.0))
        } else {
            (1.0 / (2.0 * (1.0 - rand_val))).powf(1.0 / (eta + 1.0))
        };
        g1[j] = (0.5 * ((y1 + y2) - beta_q * (y2 - y1))).max(lo).min(hi);
        g2[j] = (0.5 * ((y1 + y2) + beta_q * (y2 - y1))).max(lo).min(hi);
    }
    (Individual::new(g1), Individual::new(g2))
}

/// Crossover operator enum.
#[derive(Clone, Debug)]
pub enum CrossoverMethod {
    Uniform,
    SinglePoint,
    TwoPoint,
    Blend { alpha: f64 },
}

pub fn crossover(
    p1: &Individual,
    p2: &Individual,
    method: &CrossoverMethod,
    rng: &mut impl Rng,
) -> (Individual, Individual) {
    match method {
        CrossoverMethod::Uniform => uniform_crossover(p1, p2, rng),
        CrossoverMethod::SinglePoint => single_point_crossover(p1, p2, rng),
        CrossoverMethod::TwoPoint => two_point_crossover(p1, p2, rng),
        CrossoverMethod::Blend { alpha } => blend_crossover(p1, p2, *alpha, rng),
    }
}

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

    fn parents() -> (Individual, Individual) {
        (
            Individual::new(vec![1.0, 2.0, 3.0, 4.0, 5.0]),
            Individual::new(vec![10.0, 20.0, 30.0, 40.0, 50.0]),
        )
    }

    #[test]
    fn test_uniform_crossover_genes_from_parents() {
        let (p1, p2) = parents();
        let mut rng = rand::thread_rng();
        let (c1, _c2) = uniform_crossover(&p1, &p2, &mut rng);
        for i in 0..5 {
            assert!(
                (c1.genome[i] == p1.genome[i] || c1.genome[i] == p2.genome[i]),
                "gene must come from a parent"
            );
        }
    }

    #[test]
    fn test_single_point_produces_valid_children() {
        let (p1, p2) = parents();
        let mut rng = rand::thread_rng();
        let (c1, c2) = single_point_crossover(&p1, &p2, &mut rng);
        assert_eq!(c1.genome.len(), 5);
        assert_eq!(c2.genome.len(), 5);
    }

    #[test]
    fn test_two_point_produces_valid_children() {
        let (p1, p2) = parents();
        let mut rng = rand::thread_rng();
        let (c1, c2) = two_point_crossover(&p1, &p2, &mut rng);
        assert_eq!(c1.genome.len(), 5);
        assert_eq!(c2.genome.len(), 5);
    }

    #[test]
    fn test_blend_crossover_within_bounds() {
        let (p1, p2) = parents();
        let mut rng = rand::thread_rng();
        let alpha = 0.5;
        for _ in 0..100 {
            let (c1, _) = blend_crossover(&p1, &p2, alpha, &mut rng);
            for i in 0..5 {
                let lo = p1.genome[i].min(p2.genome[i]);
                let hi = p1.genome[i].max(p2.genome[i]);
                let range = hi - lo;
                assert!(c1.genome[i] >= lo - alpha * range - 1e-10);
                assert!(c1.genome[i] <= hi + alpha * range + 1e-10);
            }
        }
    }

    #[test]
    fn test_crossover_enum_dispatch() {
        let (p1, p2) = parents();
        let mut rng = rand::thread_rng();
        for method in &[
            CrossoverMethod::Uniform,
            CrossoverMethod::SinglePoint,
            CrossoverMethod::TwoPoint,
            CrossoverMethod::Blend { alpha: 0.5 },
        ] {
            let (c1, c2) = crossover(&p1, &p2, method, &mut rng);
            assert_eq!(c1.genome.len(), 5);
            assert_eq!(c2.genome.len(), 5);
        }
    }
}