mahf 0.1.0

A framework for modular construction and evaluation of metaheuristics.
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

//! Recombination components for Differential Evolution (DE).

use eyre::ContextCompat;
use itertools::multizip;
use rand::Rng;
use serde::{Deserialize, Serialize};

use crate::{
    component::ExecResult,
    components::Component,
    population::{AsSolutions, AsSolutionsMut},
    problems::VectorProblem,
    Problem, State,
};

/// Performs a binomial crossover, combining two individuals from two populations at the same index.
///
/// Originally proposed for, and used as recombination in [`de`].
///
/// Requires at least two populations on the stack, where the top population is modified.
///
/// Note that this crossover only has an effect if the two populations differ from each other.
///
/// [`de`]: crate::heuristics::de
///
/// # Errors
///
/// Returns an `Err` if there are less than two populations on the stack.
#[derive(Clone, Serialize, Deserialize)]
pub struct DEBinomialCrossover {
    pc: f64,
}

impl DEBinomialCrossover {
    pub fn from_params(pc: f64) -> Self {
        Self { pc }
    }

    pub fn new<P>(pc: f64) -> Box<dyn Component<P>>
    where
        P: Problem + VectorProblem<Element = f64>,
    {
        Box::new(Self::from_params(pc))
    }
}

impl<P> Component<P> for DEBinomialCrossover
where
    P: Problem + VectorProblem<Element = f64>,
{
    fn execute(&self, problem: &P, state: &mut State<P>) -> ExecResult<()> {
        let mut populations = state.populations_mut();
        let mut rng = state.random_mut();

        let mut mutations = populations
            .try_pop()
            .wrap_err("mutated individuals are missing")?;
        let bases = populations
            .get_current()
            .wrap_err("base population is missing")?;

        for (mutation, base) in multizip((mutations.as_solutions_mut(), bases.as_solutions())) {
            let index = rng.gen_range(0..problem.dimension());

            for i in 0..problem.dimension() {
                if rng.gen::<f64>() <= self.pc || i == index {
                    mutation[i] = base[i];
                }
            }
        }

        populations.push(mutations);
        Ok(())
    }
}

/// Performs a exponential crossover, combining two individuals from two populations at the same index.
///
/// Originally proposed for, and used as recombination in [`de`].
///
/// Requires at least two populations on the stack, where the top population is modified.
///
/// Note that this crossover only has an effect if the two populations differ from each other.
///
/// [`de`]: crate::heuristics::de
///
/// # Errors
///
/// Returns an `Err` if there are less than two populations on the stack.
#[derive(Clone, Serialize, Deserialize)]
pub struct DEExponentialCrossover {
    pc: f64,
}

impl DEExponentialCrossover {
    pub fn from_params(pc: f64) -> Self {
        Self { pc }
    }

    pub fn new<P>(pc: f64) -> Box<dyn Component<P>>
    where
        P: Problem + VectorProblem<Element = f64>,
    {
        Box::new(Self::from_params(pc))
    }
}

impl<P> Component<P> for DEExponentialCrossover
where
    P: Problem + VectorProblem<Element = f64>,
{
    fn execute(&self, problem: &P, state: &mut State<P>) -> ExecResult<()> {
        let mut populations = state.populations_mut();
        let mut rng = state.random_mut();

        let mut mutations = populations.pop();
        let bases = populations.current();

        for (mutation, base) in multizip((mutations.as_solutions_mut(), bases.as_solutions())) {
            let index = rng.gen_range(0..problem.dimension());
            let mut i = index;

            loop {
                mutation[i] = base[i];
                i = (i + 1) % problem.dimension();

                if rng.gen::<f64>() > self.pc || i == index {
                    break;
                }
            }
        }

        populations.push(mutations);
        Ok(())
    }
}