Examples of Chromosome

com.trimga.Chromosome
forestry.core.genetics.Chromosome
jcgp.backend.population.Chromosome
This class encapsulates a CGP chromosome.

A chromosome contains a matrix of nodes and arrays of inputs and outputs. These elements are all interconnected, and actually form the chromosome network itself. Individual nodes can be retrieved using {@code getNode()}which requires the row and column to be specified. The same works for inputs and outputs using the associated getters, in which case only the index is necessary.

In evolutionary computation it is often necessary to make copies of chromosomes; this can be accomplished in JCGP in two ways. The recommended way to do this is using {@code copyChromosome()} in {@code Population}, but alternatively it can be done by using the {@code Chromosome} copy constructor and specifying theobject to copy from, or by using the {@code copyGenes()} method.

To illustrate this, given two chromosomes, chr1 and chr2, the following code:

chr1.copyGenes(chr2);

will modify all of chr1's connections and functions to match those of chr2, without creating a new instance. In contrast,

chr1 = new Chromosome(chr2);

creates a new instance of chromosome which is identical to chr2 and assigns it to chr1, meaning any old references to chr1 that are not updated will still refer to a chromosome that is not identical to chr2. In practice, the most reliable way is to use the copy method in {@code Population}. Assuming chr1 and chr2 are indexed 1 and 2 in {@code population} respectively,

population.copyChromosome(2, 1);

will copy chr2 into chr1 without creating new instances or requiring access to the underlying chromosome array. {@code Chromosome} offers a variety of methods to compare chromosomes as well, such as {@code compareGenesTo()} and {@code compareActiveGenesTo()}. {@code Comparable} is implementedto compare fitness value, meaning {@code compareTo()} returns a value depending the relative fitnessof the compared chromosomes.

In order to set the chromosome's input values for decoding, {@code setInputs()} should be used. A few utility methods are provided in order to retrieve random elements from the chromosome, which are used internally to initialise with random connections but also externally by mutators when performing mutations. @author Eduardo Pedroni
megamek.client.bot.ga.Chromosome
no.uib.jsparklines.data.Chromosome
org.apache.commons.math3.genetics.Chromosome
Individual in a population. Chromosomes are compared based on their fitness.
The chromosomes are IMMUTABLE, and so their fitness is also immutable and therefore it can be cached. @since 2.0
org.broad.igv.feature.Chromosome
Simple representation of a chromosome. Basically a name, length, and optionally a list of cytobands. @author jrobinso
org.encog.ml.genetic.genome.Chromosome
Implements a chromosome to genetic algorithm. This is an abstract class. Other classes are provided in this book that use this base class to train neural networks or provide an answer to the traveling salesman problem. Chromosomes are made up of genes. Genomes in this genetic algorithm consist of one or more chromosomes.
org.jgap.Chromosome
Chromosomes represent potential solutions and consist of a fixed-length collection of genes. Each gene represents a discrete part of the solution. Each gene in the Chromosome may be backed by a different concrete implementation of the Gene interface, but all genes in a respective position (locus) must share the same concrete implementation across Chromosomes within a single population (genotype). In other words, gene 1 in a chromosome must share the same concrete implementation as gene 1 in all other chromosomes in the population. @author Neil Rotstan @author Klaus Meffert @since 1.0

Examples of jcgp.backend.population.Chromosome

   * 
   */
  @Test
  public void compareTest() {
    // create a clone of the chromosome, compare - should return true
    Chromosome c = new Chromosome(chromosome);
    assertTrue("Chromosomes did not match.", chromosome.compareGenesTo(c));
    assertTrue("Symmetry not obeyed.", c.compareGenesTo(chromosome));


    // create a new random chromosome, this time they should not match
    c = new Chromosome(resources);
    assertTrue("Chromosomes did match.", !chromosome.compareGenesTo(c));
  }

View Full Code Here

Examples of jcgp.backend.population.Chromosome

    resources.setRows(3);
    resources.setInputs(3);
    resources.setOutputs(2);
    resources.setLevelsBack(3);


    Chromosome c = new Chromosome(resources);


    c.getNode(0, 0).initialise(resources.getFunction(0), c.getInput(0), c.getInput(1));
    c.getNode(1, 1).initialise(resources.getFunction(0), c.getNode(0, 0), c.getInput(1));
    c.getNode(1, 2).initialise(resources.getFunction(0), c.getNode(1, 1), c.getInput(2));


    c.getOutput(0).setSource(c.getNode(0, 0));
    c.getOutput(1).setSource(c.getNode(1, 2));
    
    return c;
  }

View Full Code Here

Examples of megamek.client.bot.ga.Chromosome

        }
    }


    // now they have a hard-coded hoard metality
    protected double getFitness(int iChromIndex) {
        Chromosome chrom = this.chromosomes[iChromIndex];
        ArrayList<MoveOption> possible = new ArrayList<MoveOption>();
        for (int iGene = 0; iGene < chromosomeDim; iGene++) {
            possible.add(new MoveOption(
                    this.moves.get(iGene)[chrom.genes[iGene]]));
        }

View Full Code Here

Examples of megamek.client.bot.ga.Chromosome

        }
        return -result + (max - distance_mod);
    }


    public MoveOption getResult() {
        Chromosome r = this.chromosomes[best];
        ArrayList<MoveOption> possible = new ArrayList<MoveOption>();
        for (int iGene = 0; iGene < chromosomeDim; iGene++) {
            possible.add(new MoveOption(this.moves.get(iGene)[r.genes[iGene]]));
        }
        Object[] move_array = possible.toArray();

View Full Code Here

Examples of megamek.client.bot.ga.Chromosome

        }
        return result;
    }


    protected void doRandomMutation(int iChromIndex) {
        Chromosome c1 = this.chromosomes[iChromIndex];
        // I don't think we need to mutate an empty chromosome
        if (c1.genes.length < 1) {
            return;
        }
        int r1 = (c1.genes.length > 2) ? Compute.randomInt(c1.genes.length - 1)

View Full Code Here

Examples of megamek.client.bot.ga.Chromosome


    public double[] getDamageUtilities() {
        int iChromIndex = populationDim - 1;
        targets.clear(); // could use ArrayList and not hashtable
        double[] result = new double[target_array.size()];
        Chromosome chromArrayList = chromosomes[iChromIndex];
        // TODO should account for high heat?
        int heat_total = 0;
        if (chromArrayList.genes[chromosomeDim - 1] >= target_array.size()) {
            chromArrayList.genes[chromosomeDim - 1] = valid_target_indexes.get(
                    0).intValue();

View Full Code Here

Examples of megamek.client.bot.ga.Chromosome

     * but the highest chance of mutation, this is where we use the primary
     * target heuristic to drive convergence
     */
    @Override
    protected void doRandomMutation(int iChromIndex) {
        Chromosome c1 = chromosomes[iChromIndex];
        // skip if it's an empty chromosome
        if (c1.genes.length < 1) {
            return;
        }
        int r1 = (c1.genes.length > 2) ? Compute.randomInt(c1.genes.length - 1)

View Full Code Here

Examples of megamek.client.bot.ga.Chromosome

        // use first weapon target as primary, not smart but good enough...
        AttackOption a = attack.get(0).get(0);
        (chromosomes[0]).genes[chromosomeDim - 1] = a.target.enemy_num;


        for (int i = 1; i < populationDim; i++) {
            Chromosome cv = chromosomes[i];
            for (int iGene = 0; iGene < chromosomeDim - 1; iGene++) {
                cv.genes[iGene] = Compute.randomInt(attack.get(iGene).size());
                if (i <= attack.size()) {
                    if (iGene + 1 == i) {
                        cv.genes[iGene] = 0; // fire

View Full Code Here

Examples of no.uib.jsparklines.data.Chromosome

        int possibleCpt = 0, retainedCpt = 0;


        for (String protein : possibleProteins) {


            String description, geneName, proteinEvidenceLevel;
            Chromosome chromosome;


            try {
                description = sequenceFactory.getHeader(protein).getSimpleProteinDescription();


                // if description is not set, return the accession instead - fix for home made fasta headers
                if (description == null || description.trim().isEmpty()) {
                    description = protein;
                }


                geneName = sequenceFactory.getHeader(protein).getGeneName();
                proteinEvidenceLevel = sequenceFactory.getHeader(protein).getProteinEvidence();


                if (proteinEvidenceLevel != null) {
                    try {
                        Integer level = new Integer(proteinEvidenceLevel);
                        proteinEvidenceLevel = GenePreferences.getProteinEvidencAsString(level);
                    } catch (NumberFormatException e) {
                        // ignore
                    }
                }


                String chromosomeNumber = geneFactory.getChromosomeForGeneName(geneName);
                chromosome = new Chromosome(chromosomeNumber);


            } catch (Exception e) {
                peptideShakerGUI.catchException(e);
                description = "Error";
                geneName = "Error";

View Full Code Here

Examples of org.apache.commons.math3.genetics.Chromosome

            
            int generation = 0;
            
            @Override
            public boolean isSatisfied(Population population) {
                Chromosome fittestChromosome = population.getFittestChromosome();
                
                if (generation == 1 || generation % 10 == 0) {
                    System.out.println("Generation " + generation + ": " + fittestChromosome.toString());
                }
                generation++;


                double fitness = fittestChromosome.fitness();
                if (Precision.equals(fitness, 0.0, 1e-6)) {
                    return true;
                } else {
                    return false;
                }
            }
        };


        System.out.println("Starting evolution ...");
        
        // run the algorithm
        Population finalPopulation = ga.evolve(initial, stoppingCondition);


        // Get the end time for the simulation.
        long endTime = System.currentTimeMillis();


        // best chromosome from the final population
        Chromosome best = finalPopulation.getFittestChromosome();
        System.out.println("Generation " + ga.getGenerationsEvolved() + ": " + best.toString());
        System.out.println("Total execution time: " + (endTime - startTime) + "ms");
  }

View Full Code Here

0 1 2 3 4 5

TOP

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.