Example usage for org.apache.commons.math3.genetics Chromosome getFitness

Introduction

In this page you can find the example usage for org.apache.commons.math3.genetics Chromosome getFitness.

Prototype

public double getFitness() 

Source Link

Document

Access the fitness of this chromosome.

Usage

From source file:it.units.malelab.sse.OperationsChromosome.java

@Override
public int compareTo(Chromosome another) {
    int fitnessComparison = -Double.compare(getFitness(), another.getFitness());
    int errorRatioComparison = -Double.compare(getStats().get(Evaluator.ResultType.ERROR_RATIO),
            ((OperationsChromosome) another).getStats().get(Evaluator.ResultType.ERROR_RATIO));
    int opsComparison = -Double.compare(getStats().get(Evaluator.ResultType.AVG_OPS),
            ((OperationsChromosome) another).getStats().get(Evaluator.ResultType.AVG_OPS));
    int sizeComparison = -Double.compare(getStats().get(Evaluator.ResultType.SIZE),
            ((OperationsChromosome) another).getStats().get(Evaluator.ResultType.SIZE));
    if (fitnessComparison != 0) {
        return fitnessComparison;
    }/* w w  w .  jav a 2  s .c  o m*/
    if (errorRatioComparison != 0) {
        return errorRatioComparison;
    }
    if (opsComparison != 0) {
        return opsComparison;
    }
    return sizeComparison;
}

---

From source file:org.apache.kylin.cube.cuboid.algorithm.generic.RouletteWheelSelection.java

@Override
public ChromosomePair select(Population population) throws IllegalArgumentException {
    // create a copy of the chromosome list
    List<Chromosome> chromosomes = Lists.newArrayList(((ListPopulation) population).getChromosomes());

    double maxFitness = 0;
    double totalFitness = 0;
    for (Chromosome o : chromosomes) {
        double fitness = o.getFitness();
        totalFitness += fitness;/*from   www.ja  va 2s . c om*/
        if (fitness > maxFitness) {
            maxFitness = fitness;
        }
    }
    return new ChromosomePair(rouletteWheel(chromosomes, totalFitness),
            rouletteWheel(chromosomes, totalFitness));
}

---

From source file:org.apache.kylin.cube.cuboid.algorithm.generic.RouletteWheelSelection.java

private Chromosome rouletteWheel(final List<Chromosome> chromosomes, final double totalFitness) {
    double rnd = (GeneticAlgorithm.getRandomGenerator().nextDouble() * totalFitness);
    double runningScore = 0;
    for (Chromosome o : chromosomes) {
        if (rnd >= runningScore && rnd <= runningScore + o.getFitness()) {
            return o;
        }//from   ww w.  j  a va  2s.com
        runningScore += o.getFitness();
    }
    return null;
}

---

From source file:tmp.GACombGraphMoore.java

public static void main(String... args) {
    // to test a stochastic algorithm is hard, so this will rather be an usage example

    // initialize a new genetic algorithm   
    GeneticAlgorithm ga = new GeneticAlgorithm(new OnePointCrossover<Integer>(), CROSSOVER_RATE,
            new RandomKeyMutation(), MUTATION_RATE, new TournamentSelection(TOURNAMENT_ARITY));

    // initial population
    Population initial = randomPopulation(args);

    System.out.println("Initial population:");
    System.out.println(initial.getFittestChromosome());
    long lastime = System.currentTimeMillis();

    // stopping conditions
    StoppingCondition stopCond = new FixedGenerationCount(NUM_GENERATIONS);
    // best initial chromosome
    Chromosome bestInitial = initial.getFittestChromosome();

    // run the algorithm
    //        Population finalPopulation = ga.evolve(initial, stopCond);
    double bestfit = initial.getFittestChromosome().fitness();
    Population current = initial;/*from   www.  ja  v a2  s.  c  o  m*/
    int generationsEvolved = 0;
    //        while (!stopCond.isSatisfied(current)) {
    while (bestfit != 0.0) {
        current = ga.nextGeneration(current);
        generationsEvolved++;
        Chromosome bestFinal = current.getFittestChromosome();
        //            System.out.print(bestFinal);
        double atualfit = bestFinal.getFitness();
        if (atualfit > bestfit || System.currentTimeMillis() - lastime > HOUR) {
            bestfit = atualfit;
            //                String strbest = generationsEvolved + "-f=" + atualfit + "-" + ((MinPermutations) bestFinal).decode(sequence).toString().replaceAll(" ", "") + "\n";
            String strbest = generationsEvolved + "-f=" + atualfit;
            //                UtilTmp.dumpString(strbest);
            System.out.println(strbest);
            //                strbest = bestFinal.toString();
            //                UtilTmp.dumpString(strbest);
            System.out.println(strbest);
            //                System.out.println();
            lastime = System.currentTimeMillis();
        }
    }

    // best chromosome from the final population
    Chromosome bestFinal = current.getFittestChromosome();
    System.out.println("Best initial:");
    System.out.println(bestInitial);
    System.out.println(((MinPermutations) bestInitial).decode(sequence));
    System.out.println("Best result:");
    System.out.println(bestFinal);
    System.out.println(((MinPermutations) bestFinal).decode(sequence));

    // the only thing we can test is whether the final solution is not worse than the initial one
    // however, for some implementations of GA, this need not be true :)
    //        Assert.assertTrue(bestFinal.compareTo(bestInitial) > 0);
    //System.out.println(bestInitial);
    //System.out.println(bestFinal);
}

---

From source file:tmp.GACombPermutation.java

public static void main(String... args) {
    // to test a stochastic algorithm is hard, so this will rather be an usage example

    // initialize a new genetic algorithm
    GeneticAlgorithm ga = new GeneticAlgorithm(new OnePointCrossover<Integer>(), CROSSOVER_RATE,
            new RandomKeyMutation(), MUTATION_RATE, new TournamentSelection(TOURNAMENT_ARITY));

    // initial population
    Population initial = randomPopulation(args);
    System.out.print("Graph e-");
    System.out.print(graph.getEdgeCount());
    System.out.print(" Subgraph e-");
    System.out.println(subgraph.getEdgeCount());

    System.out.println("Initial population:");
    System.out.println(initial.getFittestChromosome());
    long lastime = System.currentTimeMillis();

    // stopping conditions
    StoppingCondition stopCond = new FixedGenerationCount(NUM_GENERATIONS);

    // best initial chromosome
    Chromosome bestInitial = initial.getFittestChromosome();

    // run the algorithm
    //        Population finalPopulation = ga.evolve(initial, stopCond);
    double bestfit = initial.getFittestChromosome().fitness();
    Population current = initial;//from   ww w. j av a 2 s  . c om
    int generationsEvolved = 0;
    //        while (!stopCond.isSatisfied(current)) {
    while (bestfit != 0.0) {
        current = ga.nextGeneration(current);
        generationsEvolved++;
        Chromosome bestFinal = current.getFittestChromosome();
        //            System.out.print(bestFinal);
        double atualfit = bestFinal.getFitness();
        if (atualfit > bestfit || System.currentTimeMillis() - lastime > UtilTmp.ALERT_HOUR) {
            lastime = System.currentTimeMillis();
            System.out.print(generationsEvolved);
            System.out.print("-");
            bestfit = atualfit;
            String strbest = bestFinal.toString() + "\n";
            UtilTmp.dumpString(strbest);
            System.out.print(strbest);
            System.out.println();
        }
    }

    // best chromosome from the final population
    Chromosome bestFinal = current.getFittestChromosome();
    System.out.println("Best result:");
    System.out.println(bestFinal);

    // the only thing we can test is whether the final solution is not worse than the initial one
    // however, for some implementations of GA, this need not be true :)
    //        Assert.assertTrue(bestFinal.compareTo(bestInitial) > 0);
    //System.out.println(bestInitial);
    //System.out.println(bestFinal);
}

---