Struct Parameters 

pub struct Parameters<T: FitnessEvaluator> {
    pub population_size: usize,
    pub genetic_code_length: usize,
    pub keep_threshold: f64,
    pub mutation_chance_percent: f64,
    pub emit_result_every: usize,
    pub fitness_evaluator: T,
}

Parameters of a simulation.

Fields

population_size: usize

Size of the population for every generation. (default = 100)

Increasing this value might improve results (because of a larger gene pool), but will drastically increase time to convergence.

genetic_code_length: usize

The length of the genetic code for all cells (default = 10)

This value will most likely need to be changed in accordance with the problem you’re trying to solve.

keep_threshold: f64

mutation_chance_percent: f64

emit_result_every: usize

How many generations between result samples (default = 1000)

Since observing results are expensive (a result contains a copy of the genetic code of the current best cell), this parameter allows fine-grained control on exactly when progress is observed. Lower values will decrease performance but increase observability, and vice-versa.

fitness_evaluator: T

The fitness evaluator to use for this simulation.

Implementations

impl<T> Parameters<T>

where T: FitnessEvaluator,

pub fn new(evaluator: T) -> Parameters<T>

pub fn with_population_size(self, population_size: usize) -> Self

pub fn with_genetic_code_length(self, length: usize) -> Self

pub fn with_keep_threshold(self, threshold: f64) -> Self

pub fn with_mutation_chance_percent(self, mutation_chance: f64) -> Self