Class to represents a Parallel Genetic Algorithm (ParGA). This algorithm runs the evolution in numberOfCores different cores to speedup the evaluation. Inherits from AbstractGA. More...

#include <ParallelGeneticAlgorithm.h>

Inheritance diagram for ParallelGeneticAlgorithm< S >:

Public Member Functions
	ParallelGeneticAlgorithm ()
	Creates a default ParallelGeneticAlgorithm. Use the ParGABuilder class instead. More...

virtual string	getName () const override
	Get the Name. More...

virtual void	run () override
	Run the parallel Algorithm. More...

int	getNumberOfCores () const
	Get the number of cores used. More...

void	setNumberOfCores (int nCores)
	Set the number of cores to use. More...

void	setProblem (shared_ptr< Problem< S >> prob) override
	Sets the problem to solve. Receives a share_ptr. More...

const Problem< S > *	getProblem () const
	Get a raw pointer to the problem to solve. More...

json	to_json () const override
	Generates and returns a JSON representation of the ParallelGeneticAlgorithm. More...

Public Member Functions inherited from AbstractGA< S >
	AbstractGA ()
	Creates a new AbstractGA which initialises all parameter to default. The operators will be set to null. More...

json	to_json () const override
	Creates and returns JSON object with the GA information. More...

virtual Front< S >	getResults () const
	Returns a front with the feasible non repeated individuals in the last population of the Genetic Algorithm. This is the final solution of the algorithm to the problem at hand. More...

void	run () override
	Runs the evolutionary process. This is the main EA method. This methods is in charge of: More...

void	setProblem (shared_ptr< Problem< S >> prob) override
	Uptades the problem to solve using a share_ptr. The problem must be single-objective. More...

void	setProblem (Problem< S > *prob) override
	Uptades the problem to solve using a raw pointer. The problem must be single-objective. More...

double	getMutationRate () const
	Gets the mutation rate. More...

void	setMutationRate (double mutationRate)
	Updates the mutation rate. More...

double	getCrossRate () const
	Gets the crossover rate. More...

void	setCrossRate (double crossRate)
	Updates the crossover rate. More...

const Mutation< S > *	getMutation () const
	Gets a pointer to the mutation operator. More...

void	setMutation (unique_ptr< Mutation< S >> mutation)
	Updates the mutation operator. Takes a unique_ptr pointing to the new Mutation operator and takes the ownership. More...

const Crossover< S > *	getCrossover () const
	Returns a raw pointer to the crossover operator. More...

void	setCrossover (unique_ptr< Crossover< S >> crossover)
	Updates the crossover operator. Takes a unique_ptr pointing to the new Crossover operator and takes the ownership. More...

const Selection< S > *	getSelection () const
	Returns a raw pointer of the selection operator. More...

void	setSelection (unique_ptr< Selection< S >> selectionOperator)
	Updates the selection operator. Takes a unique_ptr pointing to the new Selection operator and takes the ownership. More...

Public Member Functions inherited from AbstractEA< S >
	AbstractEA ()
	Construct a new Abstract EA object with default parameter values. More...

	AbstractEA (const int &maxEvals, const int &popsize)
	AbstractEA constructor which initialises the parameters maxEvalautions and populationSize. More...

	AbstractEA (unique_ptr< PopulationEvaluator< S >>, const int &maxEvals, const int &popsize)
	AbstractEA constructor which initialises the parameters maxEvalautions, the populationSize and the evaluator approach. More...

int	getPopulationSize () const
	Get the population size. More...

void	setPopulationSize (int pSize)
	Setter to update the population size. More...

double	getElapsedTime () const
	Get the elapsed time of the evolutionary process. More...

const vector< S > &	getPopulation () const
	Gets a reference of the population. More...

int	getMaxEvaluations () const
	Get the maximum number of evaluations to perform. More...

void	setMaxEvaluations (int maxEval)
	Set a new maximum number of evaluations to perform. More...

const Problem< S > *	getProblem () const
	Gets a pointer to the problem which is being solved. More...

PopulationEvaluator< S > *	getEvaluator () const
	Gets a pointer to the population evaluator system. More...

void	setEvaluator (unique_ptr< PopulationEvaluator< S >> eval)
	Updates the population evaluator system with the unique_ptr. The method takes the ownership of the object behind. More...

int	getPerformedEvaluations () const
	Get the performed evaluations. More...

void	setPerformedEvaluations (int pEvals)
	Set the Performed Evaluations. Useful for the updateProgress method. More...

void	setPopulation (const vector< S > &pop)
	Set the Population. More...

virtual Evolution< S >	getEvolution () const
	Get the Evolution data. More...

int	getPrintingInterval () const
	Get the interval of checkpoints. More...

Protected Member Functions
virtual void	initProgress ()
	Starts the progress of the parallel algorithm Sets the performedEvaluations to the number of individuals in the population. More...

virtual void	updateProgress ()
	Method which updates the evolutionary progress. This method must be implemented in the subclasses and it should perform thinks like updating the number of performed evaluations, etc.

void	runEvolution ()
	Main method of the ParallelGeneticAlgorithm. Runs the evolution of the algorithm. More...

virtual void	createInitialPopulation ()
	Creates the initial population using parallel cores. More...

virtual void	reproduction (S &, S &)
	Applies the genetic operators to the given individuals. More...

virtual void	replacement (vector< S > &offsp)

string	getID () const override
	Returns the identificator of the algorithm, this is used in the to_json method. Must be implemented in the subclasses. It should return the acronym of useful id for a particular configuration. More...

S	parallelSelection (const int &init, const int &end)
	Parallel selection operator. This methods performs a binary tournament selection in the range [int, end] of the population. Used for speedup the parallel executions. More...

void	configureEnv ()
	Configures the parallel environment. More...

Protected Member Functions inherited from AbstractGA< S >
virtual void	finishProgress ()
	Finishes the evolutionary process by computing the elapsed time. More...

virtual bool	isStoppingConditionReached ()
	Checks whether the evolutionary process has reached the maximum limit. More...

virtual void	evaluatePopulation (vector< S > &pop)
	Evaluates the entire population of individuals using the problem evaluation function. More...

virtual vector< S >	createMating ()
	Generates the mating population of individuals to be evaluated. The individuals are selected and after that the genetic operators are applied here. More...

Protected Member Functions inherited from AbstractEA< S >
virtual void	updateEvolution (vector< S > &pop)

virtual void	updateEvolution (const int &checkpoint, vector< S > &)

Protected Attributes
int	numberOfCores

int	chunks

ParallelPRNG	prng

vector< S >	individuals

shared_ptr< Problem< S > >	problem

vector< float >	bestEvolution

Protected Attributes inherited from AbstractGA< S >
double	mutationRate

double	crossRate

unique_ptr< Mutation< S > >	mutation

unique_ptr< Crossover< S > >	crossover

unique_ptr< Selection< S > >	selection

unique_ptr< Replacement< S > >	replacement

Protected Attributes inherited from AbstractEA< S >
int	maxEvaluations

int	performedEvaluations

int	populationSize

vector< S >	population

shared_ptr< Problem< S > >	problem

unique_ptr< PopulationEvaluator< S > >	evaluator

Evolution< S >	evolution

AvgEvolution< S >	avgEvolution

std::chrono::system_clock::time_point	startTime

std::chrono::system_clock::time_point	endTime

double	elapsedTime

int	nextCheckpoint

int	evolutionInterval

Additional Inherited Members
Static Public Attributes inherited from AbstractGA< S >
static double	DEFAULT_MUTATION_RATE = 0.05
	Default mutation rate for GAs set to 0.05. More...

static double	DEFAULT_CROSSOVER_RATE = 0.8
	Default crossover rate to GAs set to 0.8. More...

static int	DEFAULT_POPULATION_SIZE = 32
	Default population size to GAs set to 32 individuals. More...

Static Public Attributes inherited from AbstractEA< S >
static const int	DEFAULT_POPULATION_SIZE = 100
	Default population size equal to 100 individuals. More...

static const int	DEFAULT_EVALUATIONS_LIMIT = 100000
	Default evaluation limit equal to 100000 evaluations. More...

static const std::string	NAME = "Algorithm"

static const std::string	MAX_EVALUATIONS = "Max Evaluations"

static const std::string	POP_SIZE = "Population Size"

static const std::string	ELAPSED_TIME = "Elapsed Time"

static const std::string	EVALUATOR = "Evaluator"

static const int	EVOLUTION_SIZE = 10

Detailed Description

template<class S>
class ParallelGeneticAlgorithm< S >

Class to represents a Parallel Genetic Algorithm (ParGA). This algorithm runs the evolution in numberOfCores different cores to speedup the evaluation. Inherits from AbstractGA.

Template Parameters

S	Type of individual in the population

Constructor & Destructor Documentation

◆ ParallelGeneticAlgorithm()

template<class S >

ParallelGeneticAlgorithm< S >::ParallelGeneticAlgorithm

Creates a default ParallelGeneticAlgorithm. Use the ParGABuilder class instead.

Template Parameters

Problem
S

Member Function Documentation

◆ configureEnv()

template<class S >

void ParallelGeneticAlgorithm< S >::configureEnv

protected

Configures the parallel environment.

Template Parameters

S

◆ createInitialPopulation()

template<class S >

void ParallelGeneticAlgorithm< S >::createInitialPopulation

protectedvirtual

Creates the initial population using parallel cores.

Template Parameters

V
O

Reimplemented from AbstractGA< S >.

◆ getID()

template<class S >

string ParallelGeneticAlgorithm< S >::getID ( ) const

inlineoverrideprotectedvirtual

Returns the identificator of the algorithm, this is used in the to_json method. Must be implemented in the subclasses. It should return the acronym of useful id for a particular configuration.

Returns: string

Implements AbstractEA< S >.

◆ getName()

template<class S >

virtual string ParallelGeneticAlgorithm< S >::getName ( ) const

inlineoverridevirtual

Get the Name.

Returns: string

Implements AbstractGA< S >.

◆ getNumberOfCores()

template<class S >

int ParallelGeneticAlgorithm< S >::getNumberOfCores ( ) const

inline

Get the number of cores used.

Returns: int

◆ getProblem()

template<class S >

const Problem<S>* ParallelGeneticAlgorithm< S >::getProblem ( ) const

inline

Get a raw pointer to the problem to solve.

Returns: const Problem*

◆ initProgress()

template<class S >

void ParallelGeneticAlgorithm< S >::initProgress

protectedvirtual

Starts the progress of the parallel algorithm Sets the performedEvaluations to the number of individuals in the population.

Template Parameters

Problem
S

Reimplemented from AbstractGA< S >.

◆ parallelSelection()

template<class S >

S ParallelGeneticAlgorithm< S >::parallelSelection	(	const int &	init,
		const int &	end
	)

protected

Parallel selection operator. This methods performs a binary tournament selection in the range [int, end] of the population. Used for speedup the parallel executions.

Template Parameters

Problem
S

Parameters

init
end

Returns

◆ reproduction()

template<class S >

void ParallelGeneticAlgorithm< S >::reproduction	(	S &	child1,
		S &	child2
	)

protectedvirtual

Applies the genetic operators to the given individuals.

Template Parameters

Problem
S

Reimplemented from AbstractGA< S >.

◆ run()

template<class S >

void ParallelGeneticAlgorithm< S >::run

overridevirtual

Run the parallel Algorithm.

Starts and stops the timers
Creates the initial population
Run the evolution cycle
Finishes the progress
Template Parameters

Problem

S

Implements AbstractEA< S >.

◆ runEvolution()

template<class S >

void ParallelGeneticAlgorithm< S >::runEvolution

protected

Main method of the ParallelGeneticAlgorithm. Runs the evolution of the algorithm.

Template Parameters

Problem
S

◆ setNumberOfCores()

template<class S >

void ParallelGeneticAlgorithm< S >::setNumberOfCores ( int nCores )

inline

Set the number of cores to use.

Parameters

nCores

◆ setProblem()

template<class S >

void ParallelGeneticAlgorithm< S >::setProblem ( shared_ptr< Problem< S >> prob )

overridevirtual

Sets the problem to solve. Receives a share_ptr.

Template Parameters

S

Parameters

prob

Reimplemented from AbstractEA< S >.

◆ to_json()

template<class S >

json ParallelGeneticAlgorithm< S >::to_json

overridevirtual

Generates and returns a JSON representation of the ParallelGeneticAlgorithm.

Template Parameters

S

Returns: json

Reimplemented from AbstractEA< S >.

Member Data Documentation

◆ chunks

template<class S >

int ParallelGeneticAlgorithm< S >::chunks

protected

Chunks of population for each core

◆ numberOfCores

template<class S >

int ParallelGeneticAlgorithm< S >::numberOfCores

protected

Number of cores to run in parallel

◆ prng

template<class S >

ParallelPRNG ParallelGeneticAlgorithm< S >::prng

protected

Random number generator

The documentation for this class was generated from the following file:

include/dignea/algorithms/singleobjective/ParallelGeneticAlgorithm.h

Public Member Functions

Protected Member Functions

Protected Attributes

Additional Inherited Members

Detailed Description

template<class S> class ParallelGeneticAlgorithm< S >

Constructor & Destructor Documentation

◆ ParallelGeneticAlgorithm()

Member Function Documentation

◆ configureEnv()

◆ createInitialPopulation()

◆ getID()

◆ getName()

◆ getNumberOfCores()

◆ getProblem()

◆ initProgress()

◆ parallelSelection()

◆ reproduction()

◆ run()

◆ runEvolution()

◆ setNumberOfCores()

◆ setProblem()

◆ to_json()

Member Data Documentation

◆ chunks

◆ numberOfCores

◆ prng

template<class S>
class ParallelGeneticAlgorithm< S >