Package ec.eda.dovs

Class DOVSSpecies

java.lang.Object

ec.Species

ec.vector.VectorSpecies

ec.vector.IntegerVectorSpecies

ec.eda.dovs.DOVSSpecies

All Implemented Interfaces:

Prototype, Setup, Serializable, Cloneable

Direct Known Subclasses:

HyperboxSpecies

public class DOVSSpecies
extends IntegerVectorSpecies

DOVSSpecies is a IntegerVectorSpecies which implements DOVS algorithm. The two most important method for a Species in DOVS problem is updateMostPromisingArea(...) and mostPromisingAreaSamples(...). We call these two methods in sequence to first determine an area around best individual and sample new individual from that area. However, there are several ways to implements these two methods, thus, we let the subclasses to determine the implementation of these two method, e.g. HyperboxSpecies.

DOVSSpecies must be used in combination with DOVSBreeder, which will call it at appropriate times to reproduce new individuals for next generations. It must also be used in combination with DOVSInitializer and DOVSEvaluator. The former will be used to generate the initial population, and the later will determine a suitable number of evaluation for each individual.

Parameters

base.initial-reps Integer > 1	Base value of number of evaluations for each individual.
base.warmup Integer > 1	Number of trial we want to randomize one dimension of the individual, used for sampling.
base.constraints-size Integer	Number of constraints for the initial optimization problem. link
base.constraints-A String	A string of double number separate by whitespace specified the left hand side coefficients of the constraint Axinvalid input: '<'=b.
base.constraints-B Double	A double number specified the right hand side of the constraint Ax<=b.
base.stochastic Boolean (default = false)	Is it the problem a stochastic problem?

Default Base
dovs.species

Parameter bases

See Also:

• Serialized Form

Field Summary

Fields

Modifier and Type	Field	Description
ArrayList<double[]>	A	Constraint coefficients
ArrayList<Individual>	activeSolutions	activeSolutions contains all the samples that is on the boundary of the most promising area.
double[]	b	Constratin coefficients
ArrayList<CornerMap>	corners	CornerMaps for the all the visisted individuals.
ArrayList<Individual>	Ek	This is the Ek in original paper, where is the collection all the individuals evaluated in generation k.
int	initialReps	Base value of number evaluation for each individual.
long	numOfTotalSamples	This is for future using.
int	optimalIndex	This integer indicate the index of optimal individual in the visited array.
static final String	P_A
static final String	P_B
static final String	P_CONSTRAINTS_SIZE
static final String	P_DOVS_SPECIES
static final String	P_INITIAL_REPETITION
static final String	P_STOCHASTIC
static final String	P_WARM_UP
int	repetition	This value will be updated at each generation to determine how many evaluation is needed for one individual.
boolean	stochastic	Is the problem a stochastic problem.
ArrayList<Individual>	visited	This list contains all the sample we have visited during current algorithm run.
HashMap<Individual,Integer>	visitedIndexMap	Given a individual, return the index of this individual in ArrayList visited
int	warmUp	warm up period for RMD sampling.

Fields inherited from class ec.vector.IntegerVectorSpecies

C_RANDOM_WALK_MUTATION, C_RESET_MUTATION, maxGene, minGene, mutationIsBounded, mutationType, P_MAXGENE, P_MINGENE, P_MUTATION_BOUNDED, P_MUTATIONTYPE, P_NUM_SEGMENTS, P_RANDOM_WALK_PROBABILITY, P_SEGMENT, P_SEGMENT_END, P_SEGMENT_START, P_SEGMENT_TYPE, randomWalkProbability, V_RANDOM_WALK_MUTATION, V_RESET_MUTATION

Fields inherited from class ec.vector.VectorSpecies

C_ANY_POINT, C_GEOMETRIC, C_INTERMED_RECOMB, C_LINE_RECOMB, C_NONE, C_ONE_POINT, C_ONE_POINT_NO_NOP, C_SIMULATED_BINARY, C_TWO_POINT, C_TWO_POINT_NO_NOP, C_UNIFORM, chunksize, crossoverDistributionIndex, crossoverProbability, crossoverType, duplicateRetries, dynamicInitialSize, genomeIncreaseProbability, genomeResizeAlgorithm, genomeSize, lineDistance, maxInitialSize, minInitialSize, mutationProbability, P_CHUNKSIZE, P_CROSSOVER_DISTRIBUTION_INDEX, P_CROSSOVERPROB, P_CROSSOVERTYPE, P_DUPLICATE_RETRIES, P_GENOMESIZE, P_GEOMETRIC_PROBABILITY, P_LINEDISTANCE, P_MUTATIONPROB, P_UNIFORM_MAX, P_UNIFORM_MIN, P_VECTORSPECIES, V_ANY_POINT, V_GEOMETRIC, V_INTERMED_RECOMB, V_LINE_RECOMB, V_ONE_POINT, V_ONE_POINT_NO_NOP, V_SIMULATED_BINARY, V_TWO_POINT, V_TWO_POINT_NO_NOP, V_UNIFORM

Fields inherited from class ec.Species

f_prototype, i_prototype, P_FITNESS, P_INDIVIDUAL, P_PIPE, pipe_prototype

Constructor Summary

Constructors

Constructor	Description
DOVSSpecies()

Method Summary

Modifier and Type	Method	Description
Parameter	defaultBase()	Returns the default base for this prototype.
void	findBestSample(EvolutionState state, Subpopulation subpop)	To find the best sample for each generation, we need to go through each individual in the current population, and also best individual and individuals in actionSolutions.
ArrayList<Individual>	mostPromisingAreaSamples(EvolutionState state, int size)	Sample from the most promising area to get new generation of individual for evaluation.
void	setup(EvolutionState state, Parameter base)	The default version of setup(...) loads requested pipelines and calls setup(...) on them and normalizes their probabilities.
ArrayList<Individual>	uniqueSamples(EvolutionState state, ArrayList<Individual> candidates)	This method will take a candidate list and identify is there is redundant individual in it.
void	updateMostPromisingArea(EvolutionState state)	Define a most promising area for search of next genertion of individuals.

Methods inherited from class ec.vector.IntegerVectorSpecies

inNumericalTypeRange, inNumericalTypeRange, loadParametersForGene, maxGene, minGene, mutationIsBounded, mutationType, randomWalkProbability

Methods inherited from class ec.vector.VectorSpecies

contains, contains, contains, contains, duplicateRetries, fill, fill, fill, fill, initializeGenomeSegmentsByEndIndices, initializeGenomeSegmentsByStartIndices, mutationProbability, newIndividual, setupGenome

Methods inherited from class ec.Species

buildMisc, clone, newIndividual, newIndividual, updateIndividual, updateSubpopulation

Methods inherited from class java.lang.Object

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

Field Details

P_DOVS_SPECIES

public static final String P_DOVS_SPECIES

See Also:

• Constant Field Values

P_INITIAL_REPETITION

public static final String P_INITIAL_REPETITION

See Also:

• Constant Field Values

P_STOCHASTIC

public static final String P_STOCHASTIC

See Also:

• Constant Field Values

P_CONSTRAINTS_SIZE

public static final String P_CONSTRAINTS_SIZE

See Also:

• Constant Field Values

P_A

public static final String P_A

See Also:

• Constant Field Values

P_B

public static final String P_B

See Also:

• Constant Field Values

P_WARM_UP

public static final String P_WARM_UP

See Also:

• Constant Field Values

optimalIndex

public int optimalIndex

This integer indicate the index of optimal individual in the visited array.

warmUp

public int warmUp

warm up period for RMD sampling.

visited

public ArrayList<Individual> visited

This list contains all the sample we have visited during current algorithm run.

visitedIndexMap

public HashMap<Individual,Integer> visitedIndexMap

Given a individual, return the index of this individual in ArrayList visited

corners

public ArrayList<CornerMap> corners

CornerMaps for the all the visisted individuals. This record the key-value pair for each individuals, where key is the coordinates and value is individual itself.

activeSolutions

public ArrayList<Individual> activeSolutions

activeSolutions contains all the samples that is on the boundary of the most promising area.

Ek

public ArrayList<Individual> Ek

This is the Ek in original paper, where is the collection all the individuals evaluated in generation k.

stochastic

public boolean stochastic

Is the problem a stochastic problem.

initialReps

public int initialReps

Base value of number evaluation for each individual.

repetition

public int repetition

This value will be updated at each generation to determine how many evaluation is needed for one individual. It make use of the initialReps.

numOfTotalSamples

public long numOfTotalSamples

This is for future using.

A

public ArrayList<double[]> A

Constraint coefficients

b

public double[] b

Constratin coefficients

Constructor Details

DOVSSpecies

public DOVSSpecies()

Method Details

defaultBase

public Parameter defaultBase()

Description copied from interface: Prototype

Returns the default base for this prototype. This should generally be implemented by building off of the static base() method on the DefaultsForm object for the prototype's package. This should be callable during setup(...).

Specified by:

defaultBase in interface Prototype

Overrides:

defaultBase in class VectorSpecies

setup

public void setup(EvolutionState state,
 Parameter base)

Description copied from class: Species

The default version of setup(...) loads requested pipelines and calls setup(...) on them and normalizes their probabilities. If your individual prototype might need to know special things about the species (like parameters stored in it), then when you override this setup method, you'll need to set those parameters BEFORE you call super.setup(...), because the setup(...) code in Species sets up the prototype.

Specified by:

setup in interface Prototype

Specified by:

setup in interface Setup

Overrides:

setup in class IntegerVectorSpecies

See Also:

• Prototype.setup(EvolutionState,Parameter)

updateMostPromisingArea

public void updateMostPromisingArea(EvolutionState state)

Define a most promising area for search of next genertion of individuals.

mostPromisingAreaSamples

public ArrayList<Individual> mostPromisingAreaSamples(EvolutionState state,
 int size)

Sample from the most promising area to get new generation of individual for evaluation.

findBestSample

public void findBestSample(EvolutionState state,
 Subpopulation subpop)

To find the best sample for each generation, we need to go through each individual in the current population, and also best individual and individuals in actionSolutions. These three type of individuals are exactly the individuals evaluated in DOVSEvaluator.

uniqueSamples

public ArrayList<Individual> uniqueSamples(EvolutionState state,
 ArrayList<Individual> candidates)

This method will take a candidate list and identify is there is redundant individual in it. If yes, it will get rid of the redundant individuals. After that, it will check if all the samples from this generation have been visited in previous generation. If yes, it will retrieve the samples from previous generations.

base.species

species (the subpopulations' species)