jenes.utils.multitasking

Class Runner

java.lang.Object

jenes.utils.multitasking.Runner

Direct Known Subclasses:

MultiThreadRunner

public abstract class Runner
extends java.lang.Object

The Runner is the abstraction of an execution enviroinment per Jenes. By now this approach has been used to parallelize the fitness evaluation step. The programmer have to override the default call-backs than use execute(jenes.GeneticAlgorithm) or execute(jenes.GeneticAlgorithm, boolean) to activate Genetic Algorithm to start evolving in the enviroinment.

Since:

2.0

Field Summary

Fields

Modifier and Type	Field and Description
protected GeneticAlgorithm	algorithm The algorithm to evolve in the enviroinment

Constructor Summary

Constructors

Constructor and Description
Runner()

Method Summary

Methods

Modifier and Type	Method and Description
abstract void	evaluateIndividual(Individual individual) Call-back invoked in substitution to GeneticAlgorithm.evaluateIndividual(jenes.population.Individual)
void	execute(GeneticAlgorithm algoritm) Start evolving the algorithm given as argument restarting its state
void	execute(GeneticAlgorithm algorithm, boolean restart) Start evolving the algorithm given as parameter in this enviroinment by applying to specificated restart flag.
void	execute(GeneticAlgorithm algorithm, Population<?> initialPopulation) Start evolving the algorithm given and adopting as initial population the one passed as argument
GeneticAlgorithm	getGeneticAlgorithm() Return the Genetic Algorithm currently in execution in this enviroinment
void	onEvaluationBegin(Population pop, boolean forced) Call-back invoked soon before the Population evaluation starts using the default Fitness defined per GeneticAlgorithm
void	onEvaluationEnd() Call-back invoked soon after the evaluation phase has been performed.
void	onInit() Call-back invoked soon before GeneticAlgorithm.onInit(long)
void	setAlgorithm(GeneticAlgorithm algorithm) Set the Genetic Algorithm to the runner.
void	start(boolean reset) Call-back invoked by GeneticAlgorithm.start(boolean)
void	stop() Call-back called soon after GeneticAlgorithm.onStop(long)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

algorithm

protected GeneticAlgorithm algorithm

The algorithm to evolve in the enviroinment

Constructor Detail

Runner

public Runner()

Method Detail

execute

public final void execute(GeneticAlgorithm algorithm,
           boolean restart)

Start evolving the algorithm given as parameter in this enviroinment by applying to specificated restart flag.

Parameters:

algorithm - the algorithm to evolve

restart - restart flag to pass to GeneticAlgorithm.evolve(boolean)

execute

public final void execute(GeneticAlgorithm algoritm)

Start evolving the algorithm given as argument restarting its state

Parameters:

algoritm - the algorithm to evolve

See Also:

execute(jenes.GeneticAlgorithm, boolean)

execute

public final void execute(GeneticAlgorithm algorithm,
           Population<?> initialPopulation)

Start evolving the algorithm given and adopting as initial population the one passed as argument

Parameters:

algorithm -

initialPopulation -

getGeneticAlgorithm

public GeneticAlgorithm getGeneticAlgorithm()

Return the Genetic Algorithm currently in execution in this enviroinment

Returns:

setAlgorithm

public void setAlgorithm(GeneticAlgorithm algorithm)

Set the Genetic Algorithm to the runner. This value will be overridden when execute is called.

Parameters:

algorithm -

start

public void start(boolean reset)

Call-back invoked by GeneticAlgorithm.start(boolean)

Parameters:

reset -

onInit

public void onInit()

Call-back invoked soon before GeneticAlgorithm.onInit(long)

stop

public void stop()

Call-back called soon after GeneticAlgorithm.onStop(long)

onEvaluationBegin

public void onEvaluationBegin(Population pop,
                     boolean forced)

Call-back invoked soon before the Population evaluation starts using the default Fitness defined per GeneticAlgorithm

Parameters:

pop - the population that will be evaluated

forced - if each individual of the population will be forced to be evaluated

onEvaluationEnd

public void onEvaluationEnd()

Call-back invoked soon after the evaluation phase has been performed.
WARNING: the method is called before the elapsed time per evaluation is computed so a huge work could affect measurements.

evaluateIndividual

public abstract void evaluateIndividual(Individual individual)

Call-back invoked in substitution to GeneticAlgorithm.evaluateIndividual(jenes.population.Individual)

Parameters:

individual -