neural 0.3.0

Genetic Algorithms in Rust
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

use crate::Chromosome;
use rand::{seq::IndexedRandom, Rng};

/// Trait for selecting a subset of the population
pub trait Selection<G>: Default {
    /// Selects a subset of the population
    ///
    /// ## Arguments
    ///
    /// * `population` - The population to select from
    /// * `count` - The number of individuals to select
    /// * `rng` - The random number generator
    ///
    /// ## Returns
    ///
    /// A vector of selected individuals
    fn select<'pop, R>(
        &self,
        population: &'pop [Chromosome<G>],
        count: usize,
        rng: &mut R,
    ) -> Vec<&'pop Chromosome<G>>
    where
        R: Rng + ?Sized;
}

/// Roulette wheel selection implementation
#[derive(Debug, Default)]
pub struct RouletteWheelSelection {}

impl RouletteWheelSelection {
    /// Creates a new [`RouletteWheelSelection`]
    #[must_use]
    pub fn new() -> Self {
        Self {}
    }
}

impl<G> Selection<G> for RouletteWheelSelection {
    fn select<'pop, R>(
        &self,
        population: &'pop [Chromosome<G>],
        count: usize,
        rng: &mut R,
    ) -> Vec<&'pop Chromosome<G>>
    where
        R: Rng + ?Sized,
    {
        if population.is_empty() || count == 0 {
            return Vec::new();
        }

        let total_fitness: f64 = population.iter().map(|c| c.fitness).sum();
        let mut selected: Vec<&'pop Chromosome<G>> = Vec::with_capacity(count);

        if total_fitness <= 0.0 {
            for _ in 0..count {
                if let Some(individual) = population.choose(rng) {
                    selected.push(individual);
                }
            }
            return selected;
        }

        for _ in 0..count {
            let spin = rng.random_range(0.0..total_fitness);
            let mut current_sum = 0.0;

            for individual in population {
                current_sum += individual.fitness;

                if current_sum > spin {
                    selected.push(individual);
                    break;
                }
            }
        }

        selected
    }
}

/// Tournament selection implementation
#[derive(Debug)]
pub struct TournamentSelection {
    tournament_size: usize,
}

impl TournamentSelection {
    /// Creates a new [`TournamentSelection`]
    ///
    /// ## Arguments
    ///
    /// * `tournament_size` - The tournament size
    #[must_use]
    pub fn new(tournament_size: usize) -> Self {
        Self { tournament_size }
    }
}

impl Default for TournamentSelection {
    fn default() -> Self {
        Self::new(5)
    }
}

impl<G> Selection<G> for TournamentSelection {
    fn select<'pop, R>(
        &self,
        population: &'pop [Chromosome<G>],
        count: usize,
        rng: &mut R,
    ) -> Vec<&'pop Chromosome<G>>
    where
        R: Rng + ?Sized,
    {
        let mut selected: Vec<&'pop Chromosome<G>> = Vec::with_capacity(count);

        for _ in 0..count {
            let mut best_fitness = f64::NEG_INFINITY;
            let mut best_individual: Option<&'pop Chromosome<G>> = None;
            let current_tourney_size = self.tournament_size.min(population.len());

            for _ in 0..current_tourney_size {
                if let Some(individual) = population.choose(rng) {
                    if individual.fitness > best_fitness {
                        best_fitness = individual.fitness;
                        best_individual = Some(individual);
                    }
                }
            }

            if let Some(individual) = best_individual {
                selected.push(individual);
            } else if !population.is_empty() {
                if let Some(random_choice) = population.choose(rng) {
                    selected.push(random_choice);
                }
            }
        }

        selected
    }
}

// TODO implement rank selection
// TODO implement stochastic universal sampling selection