DistributedEvolutionaryAlgorithmsFramework
Class Deaf

java.lang.Object
  DistributedEvolutionaryAlgorithmsFramework.Deaf

public class Deaf
extends java.lang.Object

Main file of the framework

Nested Class Summary

static class

Deaf.IslandMapper Class for Hadoop framework.

Field Summary

static int	dimensions
static int	migratingPopSize

Constructor Summary

Deaf(java.lang.Class<? extends FitnessFunction> fitFunction, int D, java.lang.Class<? extends Algorithm> algorithm, int generationsPerRound, int islandCount, int populationSize, int migratingIndividuals)

Method Summary

void	addSubpopulaionsConfig(java.lang.Class<? extends Algorithm> alg, int popSize, java.util.Map<java.lang.String,java.lang.Double> subpopParameters)
void	addSubpopulaionsConfig(java.lang.Class<? extends Algorithm> alg, java.util.Map<java.lang.String,java.lang.Double> subpopParameters)
static void	log(java.lang.String msg)
Individual	run(int rounds) Runs the hadoop framework with defined settings.
void	setEmigrationMethod(Population.expelEmigrantsMethod met) Determines how emigrating individuals are chosen.
void	setHistoryPath(java.lang.String r)
void	setImmigrationMethod(Population.acceptImmigrantsMethod met) Determines how immigrating individuals are handled.
void	setMapTasks(int mapTasks) Setting for hadoop.
void	setRepeatedTest(boolean b)
void	setTopology(java.lang.Class<? extends Topology> top)
void	setVerbose(boolean verb) If verbose setting is set to true, all populations with all their individuals will be written to Logger
void	setWriteHistory(boolean b)

Methods inherited from class java.lang.Object

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

Field Detail

dimensions

public static int dimensions

migratingPopSize

public static int migratingPopSize

Constructor Detail

Deaf

public Deaf(java.lang.Class<? extends FitnessFunction> fitFunction,
            int D,
            java.lang.Class<? extends Algorithm> algorithm,
            int generationsPerRound,
            int islandCount,
            int populationSize,
            int migratingIndividuals)

Parameters:

fitFunction - FitnessFunction class implementation with optimalized fitness function.

D - Number of dimensions of defined fitness function.

algorithm - Algorithm class implementation with computational algorithm for optimalization of fitness function

generationsPerRound - Number of generations (cycles of algorithm)

islandCount - Number of islands for populations. Equals number of Hadoop Map Tasks if not set with setMapTasks().

populationSize - Default size of populations if not set specifically.

migratingIndividuals - Default number of migrating individuals if not set specifically.

Method Detail

addSubpopulaionsConfig

public void addSubpopulaionsConfig(java.lang.Class<? extends Algorithm> alg,
                                   int popSize,
                                   java.util.Map<java.lang.String,java.lang.Double> subpopParameters)

Parameters:

alg - Population specific algorith

popSize - Population specific population size

subpopParameters - Map collection of parameters for given algorithm

addSubpopulaionsConfig

public void addSubpopulaionsConfig(java.lang.Class<? extends Algorithm> alg,
                                   java.util.Map<java.lang.String,java.lang.Double> subpopParameters)

Parameters:

alg - Population specific algorith

subpopParameters - Map collection of parameters for given algorithm

setTopology

public void setTopology(java.lang.Class<? extends Topology> top)

Parameters:

top - Topology extending class defining the topology of islands (determines the destination island for migrating population)

setImmigrationMethod

public void setImmigrationMethod(Population.acceptImmigrantsMethod met)

Determines how immigrating individuals are handled.

Parameters:

met - One of Population.acceptImmigrantsMethod (REPLACE_WORST, REPLACE_BEST, REPLACE_RANDOM, NO_REPLACE).

setEmigrationMethod

public void setEmigrationMethod(Population.expelEmigrantsMethod met)

Determines how emigrating individuals are chosen.

Parameters:

met - One of Population.expelEmigrantsMethod (EXPEL_BEST, EXPEL_RANDOM).

setVerbose

public void setVerbose(boolean verb)

If verbose setting is set to true, all populations with all their individuals will be written to Logger

Parameters:

verb - true/false

setRepeatedTest

public void setRepeatedTest(boolean b)

setWriteHistory

public void setWriteHistory(boolean b)

setHistoryPath

public void setHistoryPath(java.lang.String r)

setMapTasks

public void setMapTasks(int mapTasks)

Setting for hadoop.

Parameters:

mapTasks - number of Hadoop Map Tasks

run

public Individual run(int rounds)
               throws java.io.IOException,
                      ConfigurationException

Runs the hadoop framework with defined settings.

Parameters:

rounds - Number of rounds/migrations

Returns:

Returns the best Individual from all populations

Throws:

java.io.IOException

ConfigurationException

log

public static void log(java.lang.String msg)

Parameters:

msg - Message to write to Logger with Level.INFO