Class ComposedGeneticSearch<TGenome>

A composed genetic search algorithm that combines multiple genetic search strategies.

Remarks

This class implements a composite genetic search algorithm that utilizes multiple genetic search strategies, including eliminators and a final strategy. The algorithm is configured using the [[ComposedGeneticSearchConfig]] object.

The algorithm integrates the following components, which can be customized by providing custom implementations:

A [[GeneticSearchStrategyConfig]] to define the strategy for the genetic operations.
An [[IdGeneratorInterface]] to generate unique IDs for the genomes.

Type Parameters

TGenome extends BaseGenome
The type of genome objects in the population.

Implements

GeneticSearchInterface<TGenome>

Index

Constructors

constructor

new ComposedGeneticSearch<TGenome extends BaseGenome>(
    config: ComposedGeneticSearchConfig,
    strategy: GeneticSearchStrategyConfig<TGenome>,
    idGenerator?: IdGenerator<TGenome>,
): ComposedGeneticSearch<TGenome>
Constructs a new instance of the ComposedGeneticSearch class.
Type Parameters
- TGenome extends BaseGenome
Parameters
- config: ComposedGeneticSearchConfig
  The configuration for the composed genetic search algorithm.
- strategy: GeneticSearchStrategyConfig<TGenome>
  The strategy configuration for genetic operations.
- OptionalidGenerator: IdGenerator<TGenome>
  An optional ID generator for the genomes.
Returns ComposedGeneticSearch<TGenome>
- Defined in src/genetic.ts:376

Accessors

bestGenome

get bestGenome(): TGenome
Gets the best genome from the population.

Returns TGenome
The best genome from the population.
Implementation of GeneticSearchInterface.bestGenome
- Defined in src/genetic.ts:405

`Protected`bestGenomes

get bestGenomes(): Population<TGenome>
Gets the best genomes from the eliminators.

Returns Population<TGenome>
The best genomes from the eliminators.

Remarks
This method returns the best genomes from each eliminator. The best genomes are the genomes that have the highest fitness score. The best genomes are determined by calling GeneticSearch.bestGenome on each eliminator.
- Defined in src/genetic.ts:604

cache

get cache(): PhenomeCacheInterface
Retrieves the phenome cache used by the genetic search algorithm.

Returns PhenomeCacheInterface
The phenome cache instance.
Implementation of GeneticSearchInterface.cache
- Defined in src/genetic.ts:490

generation

get generation(): number
The current generation number.

Returns number
The current generation number.
Implementation of GeneticSearchInterface.generation
- Defined in src/genetic.ts:396

partitions

get partitions(): [number, number, number]
Calculates and returns the partitions of the population for the genetic operations.

Returns [number, number, number]
A tuple containing:
- The number of genomes that will survive.
- The number of genomes that will be created by crossover.
- The number of genomes that will be created by mutation.
Implementation of GeneticSearchInterface.partitions
- Defined in src/genetic.ts:470

population

get population(): Population<TGenome>
The current population of genomes.

Returns Population<TGenome>
The current population of genomes.
Implementation of GeneticSearchInterface.population
- Defined in src/genetic.ts:414
set population(population: Population<TGenome>): void
Current population.
Parameters
- population: Population<TGenome>
Returns void
Implementation of GeneticSearchInterface.population
- Defined in src/genetic.ts:430

Methods

clearCache

clearCache(): void
Clears the cache.

Returns void
Remarks
This method clears the cache, which is used to store the phenome of the genomes. The cache is used to avoid re-calculating the phenome of the genomes if they remain unchanged.
Implementation of GeneticSearchInterface.clearCache
- Defined in src/genetic.ts:573

fit

fit(config: GeneticSearchFitConfig<TGenome>): Promise<void>
Runs the genetic search algorithm.
Parameters
- config: GeneticSearchFitConfig<TGenome>
  The configuration for the genetic search algorithm.
Returns Promise<void>
A promise that resolves when the algorithm has finished running.
Implementation of GeneticSearchInterface.fit
- Defined in src/genetic.ts:511

fitStep

fitStep(
scheduler?: SchedulerInterface<TGenome>,
): Promise<GenerationFitnessColumn>
Runs a single step of the genetic search algorithm.
Parameters
- Optionalscheduler: SchedulerInterface<TGenome>
  Optional. The scheduler to use for the genetic search algorithm.
Returns Promise<GenerationFitnessColumn>
A promise that resolves with the fitness of the best genome in the population.
Implementation of GeneticSearchInterface.fitStep
- Defined in src/genetic.ts:549

getPopulationSummary

getPopulationSummary(roundPrecision?: number): PopulationSummary
Retrieves the population summary, optionally rounding the statistics to a specified precision.
Parameters
- OptionalroundPrecision: number
  Optional. The number of decimal places to round the summary statistics to. If not provided, no rounding is applied.
Returns PopulationSummary
The population summary, with statistics rounded to the specified precision if provided.
Implementation of GeneticSearchInterface.getPopulationSummary
- Defined in src/genetic.ts:501

refreshPopulation

refreshPopulation(): void
Refreshes the population.

Returns void
Remarks
This method is used to refresh the population, which is the array of genomes that are currently being evaluated. The population is refreshed by swapping the current population with the population buffer.
Implementation of GeneticSearchInterface.refreshPopulation
- Defined in src/genetic.ts:584

setPopulation

setPopulation(population: Population<TGenome>, resetIdGenerator?: boolean): void
Sets the current population of genomes.
Parameters
- population: Population<TGenome>
  The new population of genomes.
- resetIdGenerator: boolean = true
  Whether to reset the ID generator. Defaults to true.
Returns void
Implementation of GeneticSearchInterface.setPopulation
- Defined in src/genetic.ts:440

Class ComposedGeneticSearch<TGenome>

Remarks

Type Parameters

Implements

Index

Constructors

Accessors

Methods

Constructors

constructor

Type Parameters

Parameters

Returns ComposedGeneticSearch<TGenome>

Accessors

bestGenome

Returns TGenome

ProtectedbestGenomes

Returns Population<TGenome>

Remarks

cache

Returns PhenomeCacheInterface

generation

Returns number

partitions

Returns [number, number, number]

population

Returns Population<TGenome>

Parameters

Returns void

Methods

clearCache

Returns void

Remarks

fit

Parameters

Returns Promise<void>

fitStep

Parameters

Returns Promise<GenerationFitnessColumn>

getPopulationSummary

Parameters

Returns PopulationSummary

refreshPopulation

Returns void

Remarks

setPopulation

Parameters

Returns void

Settings

On This Page

`Protected`bestGenomes