Interface GeneticSearchInterface<TGenome>

A genetic search algorithm interface.

interface GeneticSearchInterface<TGenome extends BaseGenome> {
    bestGenome: TGenome;
    cache: PhenomeCacheInterface;
    generation: number;
    partitions: [number, number, number];
    population: Population<TGenome>;
    clearCache(): void;
    fit(config: GeneticSearchFitConfig<TGenome>): Promise<void>;
    fitStep(
        scheduler?: SchedulerInterface<TGenome>,
    ): Promise<GenerationFitnessColumn>;
    getPopulationSummary(roundPrecision?: number): PopulationSummary;
    refreshPopulation(): void;
    setPopulation(
        population: Population<TGenome>,
        resetIdGenerator: boolean,
    ): void;
}

Type Parameters

  • TGenome extends BaseGenome

    The type of genome objects in the population.

Implemented by

Properties

bestGenome cache generation partitions population

Methods

clearCache fit fitStep getPopulationSummary refreshPopulation setPopulation

Properties

ReadonlybestGenome

bestGenome: TGenome

Current best genome in the population.

Readonlycache

cache: PhenomeCacheInterface

Phenome cache.

Readonlygeneration

generation: number

Current generation number.

Readonlypartitions

partitions: [number, number, number]

Partition sizes of the population.

The first element is the size of the elite population, the second element is the size of the crossover population, and the third element is the size of the mutation population.

population

population: Population<TGenome>

Current population.

Methods

clearCache

fit

fitStep

getPopulationSummary

refreshPopulation

  • refreshPopulation(): void

    Refreshes the population.

    Returns void

setPopulation

  • setPopulation(population: Population<TGenome>, resetIdGenerator: boolean): void

    Sets the current population.

    Parameters

    • population: Population<TGenome>

      new population.

    • resetIdGenerator: boolean

      Whether to reset the ID generator.

    Returns void

Settings

Member Visibility

On This Page

Properties
Methods