ec.cgp.representation
Class FloatVectorIndividual

java.lang.Object
  ec.Individual
      ec.vector.VectorIndividual
          ec.cgp.representation.VectorIndividualCGP
              ec.cgp.representation.FloatVectorIndividual

All Implemented Interfaces:

ec.Prototype, ec.Setup, java.io.Serializable, java.lang.Cloneable

public class FloatVectorIndividual
extends VectorIndividualCGP

Float-based genome representation of a Cartesian Genetic Program. Gene values are restricted to floats in the range [0,1]. During program evaluation, each float value is scaled to integers in the acceptable range that is imposed by the gene's position.

Author:

David Oranchak, doranchak@gmail.com, http://oranchak.com

See Also:

Serialized Form

Field Summary

float[]	genome the genome
static java.lang.String	P_FLOATVECTORINDIVIDUAL

Fields inherited from class ec.cgp.representation.VectorIndividualCGP

expression

Fields inherited from class ec.Individual

evaluated, EVALUATED_PREAMBLE, fitness, P_INDIVIDUAL, species

Constructor Summary

FloatVectorIndividual()

Method Summary

java.lang.Object	clone() Make a full copy of this individual.
ec.util.Parameter	defaultBase() Returns the default base for this prototype.
void	defaultCrossover(ec.EvolutionState state, int thread, ec.vector.VectorIndividual ind) Convex (or "arithmetic") crossover for real-valued genomes.
void	defaultCrossover2(ec.EvolutionState state, int thread, ec.vector.VectorIndividual ind) Any-point crossover.
void	defaultMutate(ec.EvolutionState state, int thread) Mutate the genome.
boolean	equals(java.lang.Object ind) Adapted from FloatVectorIndividual.
long	genomeLength() Return the length of the genome
java.lang.String	genotypeToString() Print to a string the genotype of the Individual in a fashion intended to be parsed in again via parseGenotype(...).
java.lang.String	genotypeToStringForHumans() Generate the human-readable text of the genotype, including the program's expression.
java.lang.Object	getGenome() Get the genome
int	hashCode() Copied from FloatVectorIndividual
static void	main(java.lang.String[] args)
static float	randGaussian(ec.util.MersenneTwisterFast rand, float mean, float std) Generate a random value with a Gaussian distribution.
void	reset(ec.EvolutionState state, int thread) Initializes the individual by randomly choosing float values uniformly from mingene to maxgene.
void	setGenome(java.lang.Object gen) Set the genome
void	setup(ec.EvolutionState state, ec.util.Parameter base) Sets up the object by reading it from the parameters stored in state, built off of the parameter base base.
static void	testGaussian()

Methods inherited from class ec.vector.VectorIndividual

join, reset, setGenomeLength, size, split

Methods inherited from class ec.Individual

parseGenotype, printIndividual, printIndividual, printIndividualForHumans, readGenotype, readIndividual, readIndividual, toString, writeGenotype, writeIndividual

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, wait, wait, wait

Field Detail

P_FLOATVECTORINDIVIDUAL

public static final java.lang.String P_FLOATVECTORINDIVIDUAL

See Also:

Constant Field Values

genome

public float[] genome

the genome

Constructor Detail

FloatVectorIndividual

public FloatVectorIndividual()

Method Detail

defaultBase

public ec.util.Parameter defaultBase()

Description copied from interface: ec.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(...).

clone

public java.lang.Object clone()

Make a full copy of this individual.

Specified by:

clone in interface ec.Prototype

Overrides:

clone in class ec.Individual

defaultMutate

public void defaultMutate(ec.EvolutionState state,
                          int thread)

Mutate the genome. Adapted from FloatVectorIndividual.

Overrides:

defaultMutate in class ec.vector.VectorIndividual

defaultCrossover2

public void defaultCrossover2(ec.EvolutionState state,
                              int thread,
                              ec.vector.VectorIndividual ind)

Any-point crossover.

defaultCrossover

public void defaultCrossover(ec.EvolutionState state,
                             int thread,
                             ec.vector.VectorIndividual ind)

Convex (or "arithmetic") crossover for real-valued genomes. It is shown to yield improved convergence for regression problems (see Clegg et. al., "A new crossover technique for Cartesian genetic programming"). But I have not yet seen convergence improvement for parity and classification problems.

Overrides:

defaultCrossover in class ec.vector.VectorIndividual

reset

public void reset(ec.EvolutionState state,
                  int thread)

Initializes the individual by randomly choosing float values uniformly from mingene to maxgene. Adapted from FloatVectorIndividual.

Specified by:

reset in class ec.vector.VectorIndividual

hashCode

public int hashCode()

Copied from FloatVectorIndividual

Specified by:

hashCode in class ec.Individual

genotypeToStringForHumans

public java.lang.String genotypeToStringForHumans()

Generate the human-readable text of the genotype, including the program's expression.

Overrides:

genotypeToStringForHumans in class ec.Individual

genotypeToString

public java.lang.String genotypeToString()

Description copied from class: ec.Individual

Print to a string the genotype of the Individual in a fashion intended to be parsed in again via parseGenotype(...). The fitness and evaluated flag should not be included. The default form simply calls toString(), which is almost certainly wrong, and you'll probably want to override this to something else.

Overrides:

genotypeToString in class ec.Individual

equals

public boolean equals(java.lang.Object ind)

Adapted from FloatVectorIndividual.

Specified by:

equals in class ec.Individual

randGaussian

public static float randGaussian(ec.util.MersenneTwisterFast rand,
                                 float mean,
                                 float std)

Generate a random value with a Gaussian distribution. Currently unused but might be useful for some problems.

Parameters:

rand - The RNG to use.

mean - The mean we want for our random values.

std - The standard deviation we want for our random values.

Returns:

the random value

getGenome

public java.lang.Object getGenome()

Get the genome

Specified by:

getGenome in class VectorIndividualCGP

setGenome

public void setGenome(java.lang.Object gen)

Set the genome

Overrides:

setGenome in class ec.vector.VectorIndividual

genomeLength

public long genomeLength()

Return the length of the genome

Overrides:

genomeLength in class ec.vector.VectorIndividual

setup

public void setup(ec.EvolutionState state,
                  ec.util.Parameter base)

Description copied from interface: ec.Prototype

Sets up the object by reading it from the parameters stored in state, built off of the parameter base base. If an ancestor implements this method, be sure to call super.setup(state,base); before you do anything else.

For prototypes, setup(...) is typically called once for the prototype instance; cloned instances do not receive the setup(...) call. setup(...) may be called more than once; the only guarantee is that it will get called at least once on an instance or some "parent" object from which it was ultimately cloned.

Specified by:

setup in interface ec.Prototype

Specified by:

setup in interface ec.Setup

Overrides:

setup in class ec.Individual

testGaussian

public static void testGaussian()

main

public static void main(java.lang.String[] args)