Package de.uka.ipd.sdq.dsexplore.opt4j.optimizer

Class MatingBayes

java.lang.Object

org.opt4j.optimizer.ea.MatingCrossoverMutate

de.uka.ipd.sdq.dsexplore.opt4j.optimizer.MatingBayes

All Implemented Interfaces:

org.opt4j.optimizer.ea.Mating

public class MatingBayes
extends org.opt4j.optimizer.ea.MatingCrossoverMutate

This class is meant to contain methods that operate on a collection of Binary strings. These methods build a Bayesian Network out of the Binary Strings and sample out new Binary Strings. Each Binary String is actually a translated DesignDecisionGenoytpe (i.e a FinalBinaryGenotype)

Field Summary

Fields inherited from class org.opt4j.optimizer.ea.MatingCrossoverMutate

copy, coupler, crossover, crossoverRate, individualFactory, mutate, mutationRate, random

Constructor Summary

Constructors

Constructor

Description

MatingBayes(org.opt4j.operator.crossover.Crossover<org.opt4j.core.Genotype> crossover, org.opt4j.operator.mutate.Mutate<org.opt4j.core.Genotype> mutate, org.opt4j.operator.copy.Copy<org.opt4j.core.Genotype> copy, org.opt4j.optimizer.ea.Coupler coupler, org.opt4j.optimizer.ea.CrossoverRate crossoverRate, org.opt4j.operator.mutate.MutationRate mutationRate, org.opt4j.common.random.Rand random, org.opt4j.core.IndividualFactory individualFactory, QMLBoundDependentTacticOperatorsManager qmlTacticManager)

Method Summary

Modifier and Type	Method	Description
TacticOperatorsManager	getHeuristicManager()
Collection<org.opt4j.core.Individual>	getOffspring(int size, Collection<org.opt4j.core.Individual> parents)	This method is similar to the getSampledGenomes method, the only difference being that this method operates on a collection of Individual objects.

Methods inherited from class org.opt4j.optimizer.ea.MatingCrossoverMutate

getOffspring, getOffspringInternal, mate

Methods inherited from class java.lang.Object

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

Constructor Detail

MatingBayes

@Inject
public MatingBayes(org.opt4j.operator.crossover.Crossover<org.opt4j.core.Genotype> crossover,
                   org.opt4j.operator.mutate.Mutate<org.opt4j.core.Genotype> mutate,
                   org.opt4j.operator.copy.Copy<org.opt4j.core.Genotype> copy,
                   org.opt4j.optimizer.ea.Coupler coupler,
                   org.opt4j.optimizer.ea.CrossoverRate crossoverRate,
                   org.opt4j.operator.mutate.MutationRate mutationRate,
                   org.opt4j.common.random.Rand random,
                   org.opt4j.core.IndividualFactory individualFactory,
                   QMLBoundDependentTacticOperatorsManager qmlTacticManager)

Method Detail

getHeuristicManager

public TacticOperatorsManager getHeuristicManager()

getOffspring

public Collection<org.opt4j.core.Individual> getOffspring(int size,
                                                          Collection<org.opt4j.core.Individual> parents)

This method is similar to the getSampledGenomes method, the only difference being that this method operates on a collection of Individual objects. The Individual objects are converted to Collection of DesignDecisionGenotype objects, which are converted to FinalBinaryGenotype objects. The getSampledGenomes is invoked on this list of FinalBinaryGenotype objects to output new solutions. These are then converted systematically back to Individual type of objects

Specified by:

getOffspring in interface org.opt4j.optimizer.ea.Mating

Overrides:

getOffspring in class org.opt4j.optimizer.ea.MatingCrossoverMutate