org.cicirello.search.evo.BasePopulation Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of chips-n-salsa Show documentation

Chips-n-Salsa is a Java library of customizable, hybridizable, iterative, parallel, stochastic, and self-adaptive local search algorithms. The library includes implementations of several stochastic local search algorithms, including simulated annealing, hill climbers, as well as constructive search algorithms such as stochastic sampling. Chips-n-Salsa now also includes genetic algorithms as well as evolutionary algorithms more generally. The library very extensively supports simulated annealing. It includes several classes for representing solutions to a variety of optimization problems. For example, the library includes a BitVector class that implements vectors of bits, as well as classes for representing solutions to problems where we are searching for an optimal vector of integers or reals. For each of the built-in representations, the library provides the most common mutation operators for generating random neighbors of candidate solutions, as well as common crossover operators for use with evolutionary algorithms. Additionally, the library provides extensive support for permutation optimization problems, including implementations of many different mutation operators for permutations, and utilizing the efficiently implemented Permutation class of the JavaPermutationTools (JPT) library. Chips-n-Salsa is customizable, making extensive use of Java's generic types, enabling using the library to optimize other types of representations beyond what is provided in the library. It is hybridizable, providing support for integrating multiple forms of local search (e.g., using a hill climber on a solution generated by simulated annealing), creating hybrid mutation operators (e.g., local search using multiple mutation operators), as well as support for running more than one type of search for the same problem concurrently using multiple threads as a form of algorithm portfolio. Chips-n-Salsa is iterative, with support for multistart metaheuristics, including implementations of several restart schedules for varying the run lengths across the restarts. It also supports parallel execution of multiple instances of the same, or different, stochastic local search algorithms for an instance of a problem to accelerate the search process. The library supports self-adaptive search in a variety of ways, such as including implementations of adaptive annealing schedules for simulated annealing, such as the Modified Lam schedule, implementations of the simpler annealing schedules but which self-tune the initial temperature and other parameters, and restart schedules that adapt to run length.

There is a newer version: 7.0.1

Show newest version

/*
 * Chips-n-Salsa: A library of parallel self-adaptive local search algorithms.
 * Copyright (C) 2002-2022 Vincent A. Cicirello
 *
 * This file is part of Chips-n-Salsa (https://chips-n-salsa.cicirello.org/).
 *
 * Chips-n-Salsa is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * Chips-n-Salsa is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see .
 */

package org.cicirello.search.evo;

import java.util.ArrayList;
import org.cicirello.search.ProgressTracker;
import org.cicirello.search.operators.Initializer;
import org.cicirello.util.Copyable;

/**
 * The nested classes are for simple populations with double-valued and int-valued fitnesses.
 *
 * @author Vincent A. Cicirello, https://www.cicirello.org/
 */
abstract class BasePopulation {

  private BasePopulation() {}

  /**
   * The Population for an evolutionary algorithm where fitness values are type double.
   *
   * @param  The type of object under optimization.
   * @author Vincent A. Cicirello, https://www.cicirello.org/
   */
  static final class DoubleFitness> extends AbstractPopulation
      implements PopulationFitnessVector.Double {

    private final Initializer initializer;
    private final SelectionOperator selection;

    private final ArrayList> pop;
    private final ArrayList> nextPop;
    private final boolean[] updated;

    private final FitnessFunction.Double f;
    private final int MU;

    private final int[] selected;

    private double bestFitness;

    /**
     * Constructs the Population.
     *
     * @param n The size of the population, which must be positive.
     * @param initializer An initializer to supply the population with a means of generating random
     *     initial population members.
     * @param f The fitness function.
     * @param selection The selection operator.
     * @param tracker A ProgressTracker.
     */
    public DoubleFitness(
        int n,
        Initializer initializer,
        FitnessFunction.Double f,
        SelectionOperator selection,
        ProgressTracker tracker) {
      super(tracker);
      if (n < 1) {
        throw new IllegalArgumentException("population size n must be positive");
      }
      if (initializer == null || f == null || selection == null || tracker == null) {
        throw new NullPointerException("passed a null object for a required parameter");
      }
      this.initializer = initializer;
      this.selection = selection;

      this.f = f;
      MU = n;

      pop = new ArrayList>(MU);
      nextPop = new ArrayList>(MU);
      selected = new int[MU];
      updated = new boolean[MU];
      bestFitness = java.lang.Double.NEGATIVE_INFINITY;
    }

    /*
     * private constructor for use by split.
     */
    private DoubleFitness(BasePopulation.DoubleFitness other) {
      super(other);

      // these are threadsafe, so just copy references
      f = other.f;
      MU = other.MU;

      // split these: not threadsafe
      initializer = other.initializer.split();
      selection = other.selection.split();

      // initialize these fresh: not threadsafe or otherwise needs its own
      pop = new ArrayList>(MU);
      nextPop = new ArrayList>(MU);
      selected = new int[MU];
      updated = new boolean[MU];
      bestFitness = java.lang.Double.NEGATIVE_INFINITY;
    }

    @Override
    public BasePopulation.DoubleFitness split() {
      return new BasePopulation.DoubleFitness(this);
    }

    @Override
    public T get(int i) {
      return nextPop.get(i).getCandidate();
    }

    @Override
    public double getFitness(int i) {
      return pop.get(i).getFitness();
    }

    @Override
    public int size() {
      // Use pop.size() rather than MU -- there is a weird, unlikely, rare edge case
      // associated with use of elitism, where pop.size() may be less than MU early in search.
      return pop.size();
    }

    @Override
    public int mutableSize() {
      return MU;
    }

    /**
     * Gets fitness of the most fit candidate solution encountered in any generation.
     *
     * @return the fitness of the most fit encountered in any generation
     */
    public double getFitnessOfMostFit() {
      return bestFitness;
    }

    @Override
    public void updateFitness(int i) {
      double fit = f.fitness(nextPop.get(i).getCandidate());
      nextPop.get(i).setFitness(fit);
      updated[i] = true;
      if (fit > bestFitness) {
        bestFitness = fit;
        setMostFit(f.getProblem().getSolutionCostPair(nextPop.get(i).getCandidate().copy()));
      }
    }

    @Override
    public void select() {
      selection.select(this, selected);
      for (int j : selected) {
        nextPop.add(pop.get(j).copy());
      }
    }

    @Override
    public void replace() {
      pop.clear();
      for (PopulationMember.DoubleFitness e : nextPop) {
        pop.add(e);
      }
      nextPop.clear();
    }

    @Override
    public void initOperators(int generations) {
      selection.init(generations);
    }

    @Override
    public void init() {
      super.init();
      bestFitness = java.lang.Double.NEGATIVE_INFINITY;
      pop.clear();
      nextPop.clear();
      T newBest = null;
      for (int i = 0; i < MU; i++) {
        T c = initializer.createCandidateSolution();
        double fit = f.fitness(c);
        pop.add(new PopulationMember.DoubleFitness(c, fit));
        if (fit > bestFitness) {
          bestFitness = fit;
          newBest = c;
        }
      }
      setMostFit(f.getProblem().getSolutionCostPair(newBest.copy()));
    }
  }

  /**
   * The Population for an evolutionary algorithm where fitness values are type int.
   *
   * @param  The type of object under optimization.
   * @author Vincent A. Cicirello, https://www.cicirello.org/
   */
  static final class IntegerFitness> extends AbstractPopulation
      implements PopulationFitnessVector.Integer {

    private final Initializer initializer;
    private final SelectionOperator selection;

    private final ArrayList> pop;
    private final ArrayList> nextPop;
    private final boolean[] updated;

    private final FitnessFunction.Integer f;
    private final int MU;

    private final int[] selected;

    private int bestFitness;

    /**
     * Constructs the Population.
     *
     * @param n The size of the population, which must be positive.
     * @param initializer An initializer to supply the population with a means of generating random
     *     initial population members.
     * @param f The fitness function.
     * @param selection The selection operator.
     * @param tracker A ProgressTracker.
     * @param numElite The number of elite population members.
     */
    public IntegerFitness(
        int n,
        Initializer initializer,
        FitnessFunction.Integer f,
        SelectionOperator selection,
        ProgressTracker tracker) {
      super(tracker);
      if (n < 1) {
        throw new IllegalArgumentException("population size n must be positive");
      }
      if (initializer == null || f == null || selection == null || tracker == null) {
        throw new NullPointerException("passed a null object for a required parameter");
      }
      this.initializer = initializer;
      this.selection = selection;

      this.f = f;
      MU = n;

      pop = new ArrayList>(MU);
      nextPop = new ArrayList>(MU);
      selected = new int[MU];
      updated = new boolean[MU];
      bestFitness = java.lang.Integer.MIN_VALUE;
    }

    /*
     * private constructor for use by split.
     */
    private IntegerFitness(BasePopulation.IntegerFitness other) {
      super(other);

      // these are threadsafe, so just copy references
      f = other.f;
      MU = other.MU;

      // split these: not threadsafe
      initializer = other.initializer.split();
      selection = other.selection.split();

      // initialize these fresh: not threadsafe or otherwise needs its own
      pop = new ArrayList>(MU);
      nextPop = new ArrayList>(MU);
      selected = new int[MU];
      updated = new boolean[MU];
      bestFitness = java.lang.Integer.MIN_VALUE;
    }

    @Override
    public BasePopulation.IntegerFitness split() {
      return new BasePopulation.IntegerFitness(this);
    }

    @Override
    public T get(int i) {
      return nextPop.get(i).getCandidate();
    }

    @Override
    public int getFitness(int i) {
      return pop.get(i).getFitness();
    }

    @Override
    public int size() {
      // Use pop.size() rather than MU -- there is a weird, unlikely, rare edge case
      // associated with use of elitism, where pop.size() may be less than MU early in search.
      return pop.size();
    }

    @Override
    public int mutableSize() {
      return MU;
    }

    /**
     * Gets fitness of the most fit candidate solution encountered in any generation.
     *
     * @return the fitness of the most fit encountered in any generation
     */
    public int getFitnessOfMostFit() {
      return bestFitness;
    }

    @Override
    public void updateFitness(int i) {
      int fit = f.fitness(nextPop.get(i).getCandidate());
      nextPop.get(i).setFitness(fit);
      updated[i] = true;
      if (fit > bestFitness) {
        bestFitness = fit;
        setMostFit(f.getProblem().getSolutionCostPair(nextPop.get(i).getCandidate().copy()));
      }
    }

    @Override
    public void select() {
      selection.select(this, selected);
      for (int j : selected) {
        nextPop.add(pop.get(j).copy());
      }
    }

    @Override
    public void replace() {
      pop.clear();
      for (PopulationMember.IntegerFitness e : nextPop) {
        pop.add(e);
      }
      nextPop.clear();
    }

    @Override
    public void initOperators(int generations) {
      selection.init(generations);
    }

    @Override
    public void init() {
      super.init();
      bestFitness = java.lang.Integer.MIN_VALUE;
      pop.clear();
      nextPop.clear();
      T newBest = null;
      for (int i = 0; i < MU; i++) {
        T c = initializer.createCandidateSolution();
        int fit = f.fitness(c);
        pop.add(new PopulationMember.IntegerFitness(c, fit));
        if (fit > bestFitness) {
          bestFitness = fit;
          newBest = c;
        }
      }
      setMostFit(f.getProblem().getSolutionCostPair(newBest.copy()));
    }
  }
}