Examples of jcgp.backend.parameters.DoubleParameter

jcgp.backend.parameters.DoubleParameter
Parameter subclass for the double type. Most of the functionality is already implemented in {@code Parameter}, leaving only construction and type definition to the subclasses.

This class contains three constructors, two of which are public. One assumes the parameter is not critical and only takes a name and initial value, while the other allows the critical flag to be set as well. The third constructor is protected and allows the monitor flag to be set as well, allowing subclasses of this class to be used as monitors. See {@link DoubleMonitor} for an exampleof this usage.

The validate method is overridden here and left blank since not all parameters actually require validation, but where validation is required this method can be anonymously overridden on an instance-to-instance basis. @author Eduardo Pedroni

    super(resources);
    setFunctionSet(new SymbolicRegressionFunctions());
    setName("Symbolic regression");
    setFileExtension(".dat");
    
    errorThreshold = new DoubleParameter(0.01, "Error threshold") {
      @Override
      public void validate(Number newValue) {
        if (newValue.doubleValue() < 0) {
          status = ParameterStatus.INVALID;
          status.setDetails("Error threshold must be a positive value.");
        } else if (newValue.doubleValue() == 0) {
          status = ParameterStatus.WARNING;
          status.setDetails("An error threshold of 0 is very rigorous and difficult to achieve.");
        } else {
          status = ParameterStatus.VALID;
        }
      }
    };
    
    perfectionThreshold = new DoubleParameter(0.000001, "Perfection threshold") {
      @Override
      public void validate(Number newValue) {
        if (newValue.doubleValue() < 0) {
          status = ParameterStatus.INVALID;
          status.setDetails("Perfection threshold must be a positive value.");

View Full Code Here

   * @param resources a reference to the experiment's resources.
   */
  public ProbabilisticMutator(Resources resources) {
    super(resources);
    
    mutationProbability = new DoubleParameter(10, "Mutation probability", false, false) {
      @Override
      public void validate(Number newValue) {
        if (newValue.doubleValue() <= 0 || newValue.doubleValue() > 100) {
          status = ParameterStatus.INVALID;
          status.setDetails("Mutation rate must be > 0 and <= 100");

View Full Code Here

   * 
   * @param resources a reference to the experiment's resources.
   */
  public PercentPointMutator(final Resources resources) {
    super(resources);
    mutationRate = new DoubleParameter(10, "Percent mutation", false, false) {
      @Override
      public void validate(Number newValue) {


        int totalGenes = (resources.nodes() * (resources.arity() + 1)) + resources.outputs();
        int mutations = (int) (newValue.doubleValue() * (double) (totalGenes / 100.0));

View Full Code Here

TOP

Related Classes of jcgp.backend.parameters.DoubleParameter

javafx.beans.property.SimpleDoubleProperty

jcgp.backend.modules.mutator.PercentPointMutator

jcgp.backend.modules.mutator.ProbabilisticMutator

jcgp.backend.modules.problem.SymbolicRegressionProblem

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.