genetic_algorithm_traits/
individual.rs

1/// A single instance in the genetic algorithm.
2/// In a TSP for example, this would be and individual route.
3pub trait Individual<'a>: Clone {
4    /// The Type of cost data this individual is compatible to compute its
5    /// fitness on.
6    type IndividualCost: 'a;
7    /// Randomly change the object and therefore it's fitness
8    /// This is a key step of the genetic algorithm.
9    ///
10    /// # Arguments
11    ///
12    /// * `prob` - The probability with which the individual will mutate.
13    fn mutate(self, prob: f32) -> Self;
14    /// The `crossover` takes two individual and combines their characteristics.
15    /// The implementation depends on the problem to be solve with genetic algorithm but
16    /// both from a performance and runtime perspective, this is one of the most important
17    /// and time-consuming methods.
18    ///
19    /// # Arguments
20    ///
21    /// * `other` - The other individual that should be `crossover`ed with.
22    ///
23    fn crossover(&self, other: &Self) -> Self;
24    /// How `fit` is your individual, e.g. how well does it solve the problem you are
25    /// trying to solve with genetic algorithms. This is the metric that is maximised, e.g.
26    /// overall individuals with a very high fitness should be found.
27    ///
28    /// # Arguments
29    ///
30    /// * `cost_data` - The data that might be needed to compute your fitness. If you use
31    /// genetic algorithm to solve a traveling salesman problem, the `cost_data` will typically
32    /// contain your distance matrix.
33    ///
34    fn fitness(&self, cost_data: &Self::IndividualCost) -> f64;
35}
genetic_algorithm_traits/individual.rs

genetic_algorithm_traits/
individual.rs
