optirustic 1.2.3

This crate moved to https://github.com/s-simoncelli/nsga-rs
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

use std::marker::PhantomData;

use rand::prelude::IndexedRandom;
use rand::RngCore;

use crate::core::{Individual, OError};
use crate::operators::{BinaryComparisonOperator, PreferredSolution};

/// A trait implementing methods to choose individuals from a population for reproduction.
pub trait Selector {
    /// Select a number of individuals from the population equal to `number_of_winners`.
    ///
    /// # Arguments
    ///
    /// * `individuals`: The individuals.
    /// * `number_of_winners`: The number of winners to select.
    ///
    /// returns: `Result<Vec<Individual>, OError>`
    fn select(
        &self,
        individuals: &[Individual],
        number_of_winners: usize,
        rng: &mut dyn RngCore,
    ) -> Result<Vec<Individual>, OError> {
        let mut winners: Vec<Individual> = Vec::new();
        for _ in 0..number_of_winners {
            winners.push(self.select_fit_individual(individuals, rng)?);
        }
        Ok(winners)
    }

    /// Select the fittest individual from the population.
    ///
    /// # Arguments
    ///
    /// * `individuals`: The list of individuals.
    /// * `rng`: The random number generator. Set this to `None`, if you are not using a seed.
    ///
    ///
    /// returns: `Result<Individual, OError>`
    fn select_fit_individual(
        &self,
        individuals: &[Individual],
        rng: &mut dyn RngCore,
    ) -> Result<Individual, OError>;
}

/// Tournament selection method between multiple competitors for choosing individuals from a
/// population for reproduction. `number_of_competitors` individuals are randomly selected from the
/// population, then the most fit becomes a parent based on the provided `fitness` function.
/// More tournaments may be run to select more individuals.
pub struct TournamentSelector<Operator: BinaryComparisonOperator> {
    /// The number of competitors in each tournament. For example, 2 to run a binary tournament.
    number_of_competitors: usize,
    /// The function to use to assess the fitness and determine which individual wins a tournament.
    _fitness_function: PhantomData<Operator>,
}

impl<Operator: BinaryComparisonOperator> TournamentSelector<Operator> {
    /// Create a new tournament.
    ///
    /// # Arguments
    ///
    /// * `number_of_competitors`: The number of competitors in the tournament. Default to 2
    ///    individuals.
    ///
    /// returns: `TournamentSelector`
    pub fn new(number_of_competitors: usize) -> Self {
        Self {
            _fitness_function: PhantomData::<Operator>,
            number_of_competitors,
        }
    }
}

impl<Operator: BinaryComparisonOperator> Selector for TournamentSelector<Operator> {
    /// Select the fittest individual from the population.
    ///
    /// # Arguments
    ///
    /// * `individuals`:The individuals with the solutions.
    /// * `rng`: The random number generator. Set this to `None`, if you are not using a seed.
    ///
    /// returns: `Result<Individual, OError>`
    fn select_fit_individual(
        &self,
        individuals: &[Individual],
        rng: &mut dyn RngCore,
    ) -> Result<Individual, OError> {
        // let population = population.lock().unwrap();
        if individuals.is_empty() {
            return Err(OError::SelectorOperator(
                "BinaryComparisonOperator".to_string(),
                "The population is empty and no individual can be selected".to_string(),
            ));
        }
        if individuals.len() < self.number_of_competitors {
            return Err(OError::SelectorOperator(
                "BinaryComparisonOperator".to_string(),
                format!("The population size ({}) is smaller than the number of competitors needed in the tournament ({})", individuals.len(), self.number_of_competitors))
            );
        }
        let mut winner = individuals.choose(rng).unwrap();

        for _ in 0..self.number_of_competitors {
            let potential_winner = individuals.choose(rng).unwrap();
            let preferred_sol = Operator::compare(winner, potential_winner)?;
            if preferred_sol == PreferredSolution::Second {
                winner = potential_winner;
            } else if preferred_sol == PreferredSolution::MutuallyPreferred {
                // randomly select winner
                winner = [winner, potential_winner].choose(rng).unwrap();
            }
        }

        Ok(winner.clone())
    }
}