org.jgap.symbolic

Class SymbolicRegression

java.lang.Object

org.jgap.gp.GPProblem

org.jgap.symbolic.SymbolicRegression

public class SymbolicRegression
extends GPProblem

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	SymbolicRegression.FormulaFitnessFunction Fitness function for evaluating the produced fomulas, represented as GP programs.

Field Summary

Fields

Modifier and Type	Field and Description
static int	adfArity
static java.lang.String[]	adfFunctions
static java.lang.String	adfType
static boolean	alldifferentVariables
static double	alldifferentVariablesPenalty
static java.util.ArrayList<java.lang.Double>	constants
static float	crossoverProb
static float	dynamizeArityProb
static long	endTime
static java.lang.String	errorMethod
static boolean	foundPerfect
static double	functionProb
static java.lang.String[]	functions
static double	hitsCriteria
static int[]	ignoreVariables
static double	lowerRange
static boolean	makeTimeSeries
static boolean	makeTimeSeriesWithIndex
static int	maxCrossoverDepth
static int	maxInitDepth
static int	maxNodes
static int	minInitDepth
static int	minNodes
static double	minNodesPenalty
static int	modReplace
static float	mutationProb
static double	newChromsPercent
static boolean	noCommandGeneCloning
static boolean	noTerminals
static int	numEvolutions
static int	numInputVariables
static int	numRows
static java.lang.Integer	outputVariable
static int	populationSize
static java.lang.String	presentation
static int	programCreationMaxTries
static float	reproductionProb
static int	resultPrecision
static java.lang.String	returnType
static double	samplePCT
static double	scaleError
static boolean	showAllGenerations
static boolean	showPopulation
static boolean	showProgression
static boolean	showResults
static boolean	showSimiliar
static java.lang.String	similiarSortMethod
static long	startTime
static double	stopCriteriaFitness
static boolean	strictProgramCreation
static boolean	terminalWholeNumbers
static java.lang.Double[][]	testData
static int	tournamentSelectorSize
static double	upperRange
static boolean	useADF
static boolean	useProgramCache
static double	validationPCT
static java.lang.String[]	variableNames
static Variable[]	variables
static boolean	verboseOutput

Constructor Summary

Constructors

Constructor and Description
SymbolicRegression(GPConfiguration a_conf)

Method Summary

Modifier and Type	Method and Description
static double[]	calcAllErrors(double error, double[] errors)
static double	correlation(java.lang.Double[] y1, double[] y2, int n) Due to lazyness this is borrowed from Weka's weka/core/Utils.java, slightly edited.
GPGenotype	create() This method is used for setting up the commands and terminals that can be used to solve the problem.
static java.lang.Double	evalData(IGPProgram ind, java.lang.Double[] data)
static double[]	getMinMax(double[] arr)
static void	main(java.lang.String[] args) Starts the example.
static void	myOutputSolution(IGPProgram a_best, int gen) Outputs the best solution until now at standard output.
static void	readFile(java.lang.String file)
static java.lang.Double[][]	transposeMatrix(java.lang.Double[][] m)
static double	validateData(IGPProgram ind)

Methods inherited from class org.jgap.gp.GPProblem

createTree, getGPConfiguration, showTree, showTree

Methods inherited from class java.lang.Object

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

Field Detail

numInputVariables

public static int numInputVariables

variables

public static Variable[] variables

variableNames

public static java.lang.String[] variableNames

outputVariable

public static java.lang.Integer outputVariable

ignoreVariables

public static int[] ignoreVariables

constants

public static java.util.ArrayList<java.lang.Double> constants

numRows

public static int numRows

foundPerfect

public static boolean foundPerfect

minInitDepth

public static int minInitDepth

maxInitDepth

public static int maxInitDepth

populationSize

public static int populationSize

maxCrossoverDepth

public static int maxCrossoverDepth

programCreationMaxTries

public static int programCreationMaxTries

numEvolutions

public static int numEvolutions

verboseOutput

public static boolean verboseOutput

maxNodes

public static int maxNodes

functionProb

public static double functionProb

reproductionProb

public static float reproductionProb

mutationProb

public static float mutationProb

crossoverProb

public static float crossoverProb

dynamizeArityProb

public static float dynamizeArityProb

newChromsPercent

public static double newChromsPercent

tournamentSelectorSize

public static int tournamentSelectorSize

noCommandGeneCloning

public static boolean noCommandGeneCloning

strictProgramCreation

public static boolean strictProgramCreation

useProgramCache

public static boolean useProgramCache

lowerRange

public static double lowerRange

upperRange

public static double upperRange

terminalWholeNumbers

public static boolean terminalWholeNumbers

returnType

public static java.lang.String returnType

presentation

public static java.lang.String presentation

adfArity

public static int adfArity

adfType

public static java.lang.String adfType

useADF

public static boolean useADF

functions

public static java.lang.String[] functions

adfFunctions

public static java.lang.String[] adfFunctions

scaleError

public static double scaleError

startTime

public static long startTime

endTime

public static long endTime

stopCriteriaFitness

public static double stopCriteriaFitness

showPopulation

public static boolean showPopulation

showSimiliar

public static boolean showSimiliar

similiarSortMethod

public static java.lang.String similiarSortMethod

showProgression

public static boolean showProgression

showAllGenerations

public static boolean showAllGenerations

showResults

public static boolean showResults

resultPrecision

public static int resultPrecision

samplePCT

public static double samplePCT

hitsCriteria

public static double hitsCriteria

validationPCT

public static double validationPCT

testData

public static java.lang.Double[][] testData

modReplace

public static int modReplace

makeTimeSeries

public static boolean makeTimeSeries

makeTimeSeriesWithIndex

public static boolean makeTimeSeriesWithIndex

errorMethod

public static java.lang.String errorMethod

noTerminals

public static boolean noTerminals

minNodes

public static int minNodes

minNodesPenalty

public static double minNodesPenalty

alldifferentVariables

public static boolean alldifferentVariables

alldifferentVariablesPenalty

public static double alldifferentVariablesPenalty

Constructor Detail

SymbolicRegression

public SymbolicRegression(GPConfiguration a_conf)
                   throws InvalidConfigurationException

Throws:

InvalidConfigurationException

Method Detail

create

public GPGenotype create()
                  throws InvalidConfigurationException

This method is used for setting up the commands and terminals that can be used to solve the problem.

Specified by:

create in class GPProblem

Returns:

GPGenotype

Throws:

InvalidConfigurationException

readFile

public static void readFile(java.lang.String file)

transposeMatrix

public static java.lang.Double[][] transposeMatrix(java.lang.Double[][] m)

main

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

Starts the example.

Throws:

java.lang.Exception

myOutputSolution

public static void myOutputSolution(IGPProgram a_best,
                                    int gen)

Outputs the best solution until now at standard output. This is stolen (and somewhat edited) from GPGenotype.outputSolution which used log4j.

Parameters:

a_best - the fittest ProgramChromosome

correlation

public static final double correlation(java.lang.Double[] y1,
                                       double[] y2,
                                       int n)

Due to lazyness this is borrowed from Weka's weka/core/Utils.java, slightly edited. [Copyright (C) 1999-2004 University of Waikato, Hamilton, New Zealand] Weka is a great machine learning/data mining system: http://www.cs.waikato.ac.nz/~ml/weka/index.html Returns the correlation coefficient of two double vectors.

Parameters:

y1 - Double vector 1

y2 - double vector 2

n - the length of two double vectors

Returns:

the correlation coefficient

validateData

public static double validateData(IGPProgram ind)

calcAllErrors

public static double[] calcAllErrors(double error,
                                     double[] errors)

evalData

public static java.lang.Double evalData(IGPProgram ind,
                                        java.lang.Double[] data)

getMinMax

public static double[] getMinMax(double[] arr)