Package org.hortonmachine.gears.utils.optimizers.particleswarm

Interface IPSFunction

  • public interface IPSFunction
    Interface for a particle swarm fitting function that is supposed to return a value to minimize.
    Author:
    Andrea Antonello (www.hydrologis.com)

    • Method Summary

      All Methods Instance Methods Abstract Methods 
      Modifier and Type Method Description
      double evaluate​(int iterationStep, int particleNum, double[] parameters, double[]... ranges)
      Evaluates the value of the fitting function and returns representing scalar.
      double getInitialGlobalBest()
      Gives the initial global best to use (ex.
      String getPostInfoString()
      Getter for information the module might want to collect and make available.
      boolean hasConverged​(double globalBest, double[] globalBestLocations, double[] previousBestLocations)
      Evaluates if the solution has converged.
      boolean isBetter​(double evaluatedValue, double consideredBest)
      Evaluates if the supplied value is better than the supplied best.
      double optimization​(double... parameters)
      apply the optimization function.
      String optimizationDescription()  

    • Method Detail

      • evaluate

        double evaluate​(int iterationStep,
                int particleNum,
                double[] parameters,
                double[]... ranges)
         throws Exception
        Evaluates the value of the fitting function and returns representing scalar.
        Parameters:
        iterationStep - the iterationstep to monitor the process.
        particleNum - the particle number to monitor the process.
        parameters - the parameters to be used in the function.
        ranges - the valid ranges for the given parameters.
        Returns:
        the calculated value.
        Throws:
        Exception

      • optimization

        double optimization​(double... parameters)
        apply the optimization function.
        Parameters:
        parameters - the parameters needed.
        Returns:
        the optimized value;

      • optimizationDescription

        String optimizationDescription()
        Returns:
        a description for the applied optimization function.

      • isBetter

        boolean isBetter​(double evaluatedValue,
                 double consideredBest)
        Evaluates if the supplied value is better than the supplied best.

        Implementations will take care of defining whether a minimum, a maximum or some other condition has to be considered.

        Parameters:
        evaluatedValue - the value to check.
        consideredBest - the best to check against.
        Returns:
        true if the evaluatedValue is considered to be better than the supplied best.

      • hasConverged

        boolean hasConverged​(double globalBest,
                     double[] globalBestLocations,
                     double[] previousBestLocations)
        Evaluates if the solution has converged.
        Parameters:
        globalBest - the current global best.
        globalBestLocations - the locations resulting from the current global best.
        previousBestLocations - the locations of the previous iteration.
        Returns:
        true if the solution can be considered as converged.

      • getInitialGlobalBest

        double getInitialGlobalBest()
        Gives the initial global best to use (ex. for initial swarm creation).
        Returns:
        the initial global best to use.

      • getPostInfoString

        String getPostInfoString()
        Getter for information the module might want to collect and make available.
        Returns:
        the info string or null.