001package org.cpsolver.ifs.algorithms;
002
003import org.cpsolver.ifs.assignment.Assignment;
004import org.cpsolver.ifs.heuristics.NeighbourSelection;
005import org.cpsolver.ifs.model.Model;
006import org.cpsolver.ifs.model.Neighbour;
007import org.cpsolver.ifs.model.Value;
008import org.cpsolver.ifs.model.Variable;
009import org.cpsolver.ifs.solution.Solution;
010import org.cpsolver.ifs.solver.Solver;
011import org.cpsolver.ifs.util.DataProperties;
012
013/**
014 * Hill climber. In each iteration, one of the given neighbourhoods is selected first,
015 * then a neighbour is generated and it is accepted if its value
016 * {@link Neighbour#value(Assignment)} is below or equal to zero. The search is
017 * stopped after a given amount of idle iterations ( can be defined by problem
018 * property HillClimber.MaxIdle). <br>
019 * <br>
020 * Custom neighbours can be set using HillClimber.Neighbours property that should
021 * contain semicolon separated list of {@link NeighbourSelection}. By default, 
022 * each neighbour selection is selected with the same probability (each has 1 point in
023 * a roulette wheel selection). It can be changed by adding &nbsp;@n at the end
024 * of the name of the class, for example:
025 * <pre><code>
026 * HillClimber.Neighbours=org.cpsolver.ifs.algorithms.neighbourhoods.RandomMove;org.cpsolver.ifs.algorithms.neighbourhoods.RandomSwapMove@0.1
027 * </code></pre>
028 * Selector RandomSwapMove is 10&times; less probable to be selected than other selectors.
029 * When HillClimber.Random is true, all selectors are selected with the same probability, ignoring these weights.
030 * <br><br>
031 * When HillClimber.Update is true, {@link NeighbourSelector#update(Assignment, Neighbour, long)} is called 
032 * after each iteration (on the selector that was used) and roulette wheel selection 
033 * that is using {@link NeighbourSelector#getPoints()} is used to pick a selector in each iteration. 
034 * See {@link NeighbourSelector} for more details. 
035 * <br>
036 * 
037 * @version IFS 1.3 (Iterative Forward Search)<br>
038 *          Copyright (C) 2014 Tomas Muller<br>
039 *          <a href="mailto:muller@unitime.org">muller@unitime.org</a><br>
040 *          <a href="http://muller.unitime.org">http://muller.unitime.org</a><br>
041 * <br>
042 *          This library is free software; you can redistribute it and/or modify
043 *          it under the terms of the GNU Lesser General Public License as
044 *          published by the Free Software Foundation; either version 3 of the
045 *          License, or (at your option) any later version. <br>
046 * <br>
047 *          This library is distributed in the hope that it will be useful, but
048 *          WITHOUT ANY WARRANTY; without even the implied warranty of
049 *          MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
050 *          Lesser General Public License for more details. <br>
051 * <br>
052 *          You should have received a copy of the GNU Lesser General Public
053 *          License along with this library; if not see
054 *          <a href='http://www.gnu.org/licenses/'>http://www.gnu.org/licenses/</a>.
055 * @param <V> Variable
056 * @param <T> Value
057 */
058public class HillClimber<V extends Variable<V, T>, T extends Value<V, T>> extends NeighbourSearch<V, T> {
059    protected int iMaxIdleIters = 10000;
060    protected boolean iSetHCMode = false;
061
062    /**
063     * Constructor
064     * <ul>
065     * <li>HillClimber.MaxIdle ... maximum number of idle iterations (default is 200000)
066     * <li>HillClimber.Neighbours ... semicolon separated list of classes implementing {@link NeighbourSelection}
067     * <li>HillClimber.AdditionalNeighbours ... semicolon separated list of classes implementing {@link NeighbourSelection}
068     * <li>HillClimber.Random ... when true, a neighbour selector is selected randomly
069     * <li>HillClimber.Update ... when true, a neighbour selector is selected using {@link NeighbourSelector#getPoints()} weights (roulette wheel selection)
070     * </ul>
071     * @param properties solver configuration
072     */
073    public HillClimber(DataProperties properties) {
074        super(properties);
075        iMaxIdleIters = properties.getPropertyInt(getParameterBaseName() + ".MaxIdle", iMaxIdleIters);
076        iSetHCMode = properties.getPropertyBoolean(getParameterBaseName() + ".SetHCMode", iSetHCMode);
077    }
078    
079    /**
080     * Set progress phase name
081     * @param phase name of the phase in which the hill climber is used (for logging purposes)
082     */
083    public void setPhase(String phase) {
084        iPhase = phase;
085    }
086
087    /**
088     * Initialization
089     */
090    @Override
091    public void init(Solver<V, T> solver) {
092        super.init(solver);
093        if (iSetHCMode)
094            setHCMode(true);
095    }
096    
097    /**
098     * All parameters start with HillClimber base name, e.g., HillClimber.MaxIdle 
099     */
100    @Override
101    public String getParameterBaseName() { return "HillClimber"; }
102    
103    @Override
104    public NeighbourSearchContext createAssignmentContext(Assignment<V, T> assignment) {
105        return new HillClimberContext();
106    }
107    
108    public class HillClimberContext extends NeighbourSearchContext {
109        protected int iLastImprovingIter = 0;
110
111        /**
112         * Increase iteration counter
113         */
114        @Override
115        protected void incIteration(Solution<V, T> solution) {
116            super.incIteration(solution);
117            if (iIter % 10000 == 0) {
118                info("Iter=" + (iIter / 1000)+"k, NonImpIter=" + iDF2.format((iIter-iLastImprovingIter)/1000.0)+"k, Speed="+iDF2.format(1000.0*iIter/getTimeMillis())+" it/s");
119                logNeibourStatus();
120            }
121            setProgress(Math.round(100.0 * (iIter - iLastImprovingIter) / iMaxIdleIters));
122        }
123        
124        /**
125         * Stop the search after a given number of idle (not improving) iterations
126         */
127        @Override
128        protected boolean canContinue(Solution<V, T> solution) {
129            return iIter - iLastImprovingIter < iMaxIdleIters;
130        }
131        
132        /**
133         * Reset the idle iterations counter
134         */
135        @Override
136        protected void activate(Solution<V, T> solution) {
137            super.activate(solution);
138            iLastImprovingIter = iIter;
139        }
140        
141        /**
142         * Accept any move that does not worsen the solution (value &lt;= 0)
143         */
144        @Override
145        protected boolean accept(Assignment<V, T> assignment, Model<V, T> model, Neighbour<V, T> neighbour, double value, boolean lazy) {
146            if (value < 0) iLastImprovingIter = iIter;
147            return value <= 0;
148        }
149    }
150}