Clover coverage report - JGAP 3.1
Coverage timestamp: Mo Dez 11 2006 21:16:18 CET
Overview   Package   File
    FRAMES   NO FRAMES  
file stats: LOC: 225   Methods: 12
NCLOC: 123   Classes: 3
 
 Source file Conditionals Statements Methods TOTAL
TournamentSelector.java 96,4% 98,2% 100% 97,9%
coverage coverage
 1    /*
 2    * This file is part of JGAP.
 3    *
 4    * JGAP offers a dual license model containing the LGPL as well as the MPL.
 5    *
 6    * For licencing information please see the file license.txt included with JGAP
 7    * or have a look at the top of class org.jgap.Chromosome which representatively
 8    * includes the JGAP license policy applicable for any file delivered with JGAP.
 9    */
 10    package org.jgap.impl;
 11   
 12    import java.util.*;
 13    import org.jgap.*;
 14   
 15    /**
 16    * Implementation of a NaturalSelector that plays tournaments to determine
 17    * the chromosomes to be taken to the next generation.
 18    * <p>
 19    * The tournament size can be adjusted as well as the probability for selecting
 20    * an individual.
 21    *
 22    * @author Klaus Meffert
 23    * @since 2.0
 24    */
 25    public class TournamentSelector
 26    extends NaturalSelector {
 27    /** String containing the CVS revision. Read out via reflection!*/
 28    private final static String CVS_REVISION = "$Revision: 1.19 $";
 29   
 30    private TournamentSelectorConfigurable m_config
 31    = new TournamentSelectorConfigurable();
 32   
 33    private List m_chromosomes;
 34   
 35    /**
 36    * Comparator that is only concerned about fitness values
 37    */
 38    private FitnessValueComparator m_fitnessValueComparator;
 39   
 40    /**
 41    * Default constructor.<p>
 42    * Attention: The configuration used is the one set with the static method
 43    * Genotype.setConfiguration.
 44    *
 45    * @author Siddhartha Azad
 46    * @author Klaus Meffert
 47    */
 48  2 public TournamentSelector() {
 49  2 super(Genotype.getStaticConfiguration());
 50  2 init();
 51    }
 52   
 53  24 private void init() {
 54  24 m_chromosomes = new Vector();
 55  24 m_fitnessValueComparator = new FitnessValueComparator();
 56    }
 57   
 58    /**
 59    * @param a_config the configuration to use
 60    * @param a_tournament_size the size of each tournament to play
 61    * @param a_probability probability for selecting the best individuals
 62    *
 63    * @author Klaus Meffert
 64    * @since 2.0
 65    */
 66  22 public TournamentSelector(final Configuration a_config,
 67    final int a_tournament_size,
 68    final double a_probability) {
 69  22 super(a_config);
 70  22 init();
 71  22 if (a_tournament_size < 1) {
 72  2 throw new IllegalArgumentException("Tournament size must be at least 1!");
 73    }
 74  20 if (a_probability <= 0.0d || a_probability > 1.0d) {
 75  2 throw new IllegalArgumentException("Probability must be greater 0.0 and"
 76    + " less or equal than 1.0!");
 77    }
 78  18 m_config.m_tournament_size = a_tournament_size;
 79  18 m_config.m_probability = a_probability;
 80    }
 81   
 82  2 public void setTournamentSize(final int a_tournament_size) {
 83  2 if (a_tournament_size < 1) {
 84  1 throw new IllegalArgumentException("Tournament size must be at least 1!");
 85    }
 86  1 m_config.m_tournament_size = a_tournament_size;
 87    }
 88   
 89  1 public int getTournamentSize() {
 90  1 return m_config.m_tournament_size;
 91    }
 92   
 93  1 public double getProbability() {
 94  1 return m_config.m_probability;
 95    }
 96   
 97  2 public void setProbability(final double a_probability) {
 98  2 if (a_probability <= 0.0d || a_probability > 1.0d) {
 99  1 throw new IllegalArgumentException("Probability must be greater 0.0 and"
 100    + " less or equal than 1.0!");
 101    }
 102  1 m_config.m_probability = a_probability;
 103    }
 104   
 105    /**
 106    * Select a given number of Chromosomes from the pool that will move on
 107    * to the next generation population. This selection will be guided by the
 108    * fitness values. The chromosomes with the best fitness value win.
 109    *
 110    * @param a_howManyToSelect int
 111    * @param a_from_pop the population the Chromosomes will be selected from
 112    * @param a_to_pop the population the Chromosomes will be added to
 113    *
 114    * @author Klaus Meffert
 115    * @since 2.0
 116    */
 117  13 public void select(final int a_howManyToSelect, final Population a_from_pop,
 118    Population a_to_pop) {
 119  13 if (a_from_pop != null) {
 120  2 for (int i = 0; i < a_from_pop.size(); i++) {
 121  6 add(a_from_pop.getChromosome(i));
 122    }
 123    }
 124  13 List tournament = new Vector();
 125  13 RandomGenerator rn = getConfiguration().getRandomGenerator();
 126  13 int size = m_chromosomes.size();
 127  13 if (size == 0) {
 128  1 return;
 129    }
 130  12 int k;
 131  12 for (int i = 0; i < a_howManyToSelect; i++) {
 132    // choose [tournament size] individuals from the population at random
 133  78 tournament.clear();
 134  78 for (int j = 0; j < m_config.m_tournament_size; j++) {
 135  226 k = rn.nextInt(size);
 136  226 tournament.add(m_chromosomes.get(k));
 137    }
 138  78 Collections.sort(tournament, m_fitnessValueComparator);
 139  78 double prob = rn.nextDouble();
 140  78 double probAccumulated = m_config.m_probability;
 141  78 int index = 0;
 142    //play the tournament
 143  78 if (m_config.m_tournament_size > 1) {
 144  48 do {
 145  50 if (prob <= probAccumulated) {
 146  47 break;
 147    }
 148    else {
 149  3 probAccumulated += probAccumulated * (1 - m_config.m_probability);
 150  3 index++;
 151    }
 152    }
 153  3 while (index < m_config.m_tournament_size - 1);
 154    }
 155  78 a_to_pop.addChromosome( (IChromosome) tournament.get(index));
 156    }
 157    }
 158   
 159    /**
 160    * @return false as we allow to return the same chromosome multiple times
 161    *
 162    * @author Klaus Meffert
 163    * @since 2.0
 164    */
 165  1 public boolean returnsUniqueChromosomes() {
 166  1 return false;
 167    }
 168   
 169  3 public void empty() {
 170  3 m_chromosomes.clear();
 171    }
 172   
 173    /**
 174    *
 175    * @param a_chromosomeToAdd the Chromosome to add
 176    *
 177    * @author Klaus Meffert
 178    * @since 2.0
 179    */
 180  25 protected void add(final IChromosome a_chromosomeToAdd) {
 181  25 m_chromosomes.add(a_chromosomeToAdd);
 182    }
 183   
 184    /**
 185    * Comparator regarding only the fitness value. Best fitness value will
 186    * be on first position of resulting sorted list
 187    *
 188    * @author Klaus Meffert
 189    * @since 2.0
 190    */
 191    private class FitnessValueComparator
 192    implements Comparator {
 193  148 public int compare(final Object a_first, final Object a_second) {
 194  148 IChromosome chrom1 = (IChromosome) a_first;
 195  148 IChromosome chrom2 = (IChromosome) a_second;
 196  148 if (getConfiguration().getFitnessEvaluator().isFitter(chrom2.
 197    getFitnessValue(), chrom1.getFitnessValue())) {
 198  0 return 1;
 199    }
 200  148 else if (getConfiguration().getFitnessEvaluator().isFitter(
 201    chrom1.getFitnessValue(), chrom2.getFitnessValue())) {
 202  1 return -1;
 203    }
 204    else {
 205  147 return 0;
 206    }
 207    }
 208    }
 209   
 210    class TournamentSelectorConfigurable {
 211    /**
 212    * The probability for selecting the best chromosome in a tournament.
 213    * For the second-best chromosome it would be p * (1 - p).
 214    * For the third-best chromosome it would be p * (p * (1 - p)) and so forth.
 215    */
 216    public double m_probability;
 217   
 218    /**
 219    * Size of a tournament to be played, i.e. number of chromosomes taken into
 220    * account for one selection round
 221    */
 222    public int m_tournament_size;
 223   
 224    }
 225    }
 
CloverReport generated by Clover Code Coverage v1.3.5
Mo Dez 11 2006 21:17:06 CET.		   Open Source License registered to the JGAP Project. This license of Clover is provided to support the development of JGAP only.