1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
4 | * and the BEACON Center for the Study of Evolution in Action.
|
---|
5 | *
|
---|
6 | * This file is part of HeuristicLab.
|
---|
7 | *
|
---|
8 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
9 | * it under the terms of the GNU General Public License as published by
|
---|
10 | * the Free Software Foundation, either version 3 of the License, or
|
---|
11 | * (at your option) any later version.
|
---|
12 | *
|
---|
13 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
14 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
15 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
16 | * GNU General Public License for more details.
|
---|
17 | *
|
---|
18 | * You should have received a copy of the GNU General Public License
|
---|
19 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
20 | */
|
---|
21 | #endregion
|
---|
22 |
|
---|
23 | using System;
|
---|
24 | using System.Collections.Generic;
|
---|
25 | using System.Threading;
|
---|
26 | using HeuristicLab.Analysis;
|
---|
27 | using HeuristicLab.Common;
|
---|
28 | using HeuristicLab.Core;
|
---|
29 | using HeuristicLab.Data;
|
---|
30 | using HeuristicLab.Encodings.BinaryVectorEncoding;
|
---|
31 | using HeuristicLab.Optimization;
|
---|
32 | using HeuristicLab.Parameters;
|
---|
33 | using HeuristicLab.Persistence;
|
---|
34 | using HeuristicLab.Problems.Binary;
|
---|
35 | using HeuristicLab.Random;
|
---|
36 |
|
---|
37 | namespace HeuristicLab.Algorithms.ParameterlessPopulationPyramid {
|
---|
38 | // This code is based off the publication
|
---|
39 | // B. W. Goldman and W. F. Punch, "Parameter-less Population Pyramid," GECCO, pp. 785–792, 2014
|
---|
40 | // and the original source code in C++11 available from: https://github.com/brianwgoldman/Parameter-less_Population_Pyramid
|
---|
41 | [Item("Parameter-less Population Pyramid (P3)", "Binary value optimization algorithm which requires no configuration. B. W. Goldman and W. F. Punch, Parameter-less Population Pyramid, GECCO, pp. 785–792, 2014")]
|
---|
42 | [StorableType("7588f673-a2be-4bc0-a7b2-1991613a32da")]
|
---|
43 | [Creatable(CreatableAttribute.Categories.PopulationBasedAlgorithms, Priority = 400)]
|
---|
44 | public class ParameterlessPopulationPyramid : BasicAlgorithm {
|
---|
45 | public override Type ProblemType {
|
---|
46 | get { return typeof(BinaryProblem); }
|
---|
47 | }
|
---|
48 | public new BinaryProblem Problem {
|
---|
49 | get { return (BinaryProblem)base.Problem; }
|
---|
50 | set { base.Problem = value; }
|
---|
51 | }
|
---|
52 |
|
---|
53 | private readonly IRandom random = new MersenneTwister();
|
---|
54 | private List<Population> pyramid;
|
---|
55 | private EvaluationTracker tracker;
|
---|
56 |
|
---|
57 | // Tracks all solutions in Pyramid for quick membership checks
|
---|
58 | private readonly HashSet<BinaryVector> seen = new HashSet<BinaryVector>(new EnumerableBoolEqualityComparer());
|
---|
59 |
|
---|
60 | #region ParameterNames
|
---|
61 | private const string MaximumIterationsParameterName = "Maximum Iterations";
|
---|
62 | private const string MaximumEvaluationsParameterName = "Maximum Evaluations";
|
---|
63 | private const string MaximumRuntimeParameterName = "Maximum Runtime";
|
---|
64 | private const string SeedParameterName = "Seed";
|
---|
65 | private const string SetSeedRandomlyParameterName = "SetSeedRandomly";
|
---|
66 | #endregion
|
---|
67 |
|
---|
68 | #region ParameterProperties
|
---|
69 | public IFixedValueParameter<IntValue> MaximumIterationsParameter {
|
---|
70 | get { return (IFixedValueParameter<IntValue>)Parameters[MaximumIterationsParameterName]; }
|
---|
71 | }
|
---|
72 | public IFixedValueParameter<IntValue> MaximumEvaluationsParameter {
|
---|
73 | get { return (IFixedValueParameter<IntValue>)Parameters[MaximumEvaluationsParameterName]; }
|
---|
74 | }
|
---|
75 | public IFixedValueParameter<IntValue> MaximumRuntimeParameter {
|
---|
76 | get { return (IFixedValueParameter<IntValue>)Parameters[MaximumRuntimeParameterName]; }
|
---|
77 | }
|
---|
78 | public IFixedValueParameter<IntValue> SeedParameter {
|
---|
79 | get { return (IFixedValueParameter<IntValue>)Parameters[SeedParameterName]; }
|
---|
80 | }
|
---|
81 | public FixedValueParameter<BoolValue> SetSeedRandomlyParameter {
|
---|
82 | get { return (FixedValueParameter<BoolValue>)Parameters[SetSeedRandomlyParameterName]; }
|
---|
83 | }
|
---|
84 | #endregion
|
---|
85 |
|
---|
86 | #region Properties
|
---|
87 | public int MaximumIterations {
|
---|
88 | get { return MaximumIterationsParameter.Value.Value; }
|
---|
89 | set { MaximumIterationsParameter.Value.Value = value; }
|
---|
90 | }
|
---|
91 | public int MaximumEvaluations {
|
---|
92 | get { return MaximumEvaluationsParameter.Value.Value; }
|
---|
93 | set { MaximumEvaluationsParameter.Value.Value = value; }
|
---|
94 | }
|
---|
95 | public int MaximumRuntime {
|
---|
96 | get { return MaximumRuntimeParameter.Value.Value; }
|
---|
97 | set { MaximumRuntimeParameter.Value.Value = value; }
|
---|
98 | }
|
---|
99 | public int Seed {
|
---|
100 | get { return SeedParameter.Value.Value; }
|
---|
101 | set { SeedParameter.Value.Value = value; }
|
---|
102 | }
|
---|
103 | public bool SetSeedRandomly {
|
---|
104 | get { return SetSeedRandomlyParameter.Value.Value; }
|
---|
105 | set { SetSeedRandomlyParameter.Value.Value = value; }
|
---|
106 | }
|
---|
107 | #endregion
|
---|
108 |
|
---|
109 | #region ResultsProperties
|
---|
110 | private double ResultsBestQuality {
|
---|
111 | get { return ((DoubleValue)Results["Best Quality"].Value).Value; }
|
---|
112 | set { ((DoubleValue)Results["Best Quality"].Value).Value = value; }
|
---|
113 | }
|
---|
114 |
|
---|
115 | private BinaryVector ResultsBestSolution {
|
---|
116 | get { return (BinaryVector)Results["Best Solution"].Value; }
|
---|
117 | set { Results["Best Solution"].Value = value; }
|
---|
118 | }
|
---|
119 |
|
---|
120 | private int ResultsBestFoundOnEvaluation {
|
---|
121 | get { return ((IntValue)Results["Evaluation Best Solution Was Found"].Value).Value; }
|
---|
122 | set { ((IntValue)Results["Evaluation Best Solution Was Found"].Value).Value = value; }
|
---|
123 | }
|
---|
124 |
|
---|
125 | private int ResultsEvaluations {
|
---|
126 | get { return ((IntValue)Results["Evaluations"].Value).Value; }
|
---|
127 | set { ((IntValue)Results["Evaluations"].Value).Value = value; }
|
---|
128 | }
|
---|
129 | private int ResultsIterations {
|
---|
130 | get { return ((IntValue)Results["Iterations"].Value).Value; }
|
---|
131 | set { ((IntValue)Results["Iterations"].Value).Value = value; }
|
---|
132 | }
|
---|
133 |
|
---|
134 | private DataTable ResultsQualities {
|
---|
135 | get { return ((DataTable)Results["Qualities"].Value); }
|
---|
136 | }
|
---|
137 | private DataRow ResultsQualitiesBest {
|
---|
138 | get { return ResultsQualities.Rows["Best Quality"]; }
|
---|
139 | }
|
---|
140 |
|
---|
141 | private DataRow ResultsQualitiesIteration {
|
---|
142 | get { return ResultsQualities.Rows["Iteration Quality"]; }
|
---|
143 | }
|
---|
144 |
|
---|
145 |
|
---|
146 | private DataRow ResultsLevels {
|
---|
147 | get { return ((DataTable)Results["Pyramid Levels"].Value).Rows["Levels"]; }
|
---|
148 | }
|
---|
149 |
|
---|
150 | private DataRow ResultsSolutions {
|
---|
151 | get { return ((DataTable)Results["Stored Solutions"].Value).Rows["Solutions"]; }
|
---|
152 | }
|
---|
153 | #endregion
|
---|
154 |
|
---|
155 | public override bool SupportsPause { get { return true; } }
|
---|
156 |
|
---|
157 | [StorableConstructor]
|
---|
158 | protected ParameterlessPopulationPyramid(StorableConstructorFlag deserializing) : base(deserializing) { }
|
---|
159 |
|
---|
160 | protected ParameterlessPopulationPyramid(ParameterlessPopulationPyramid original, Cloner cloner)
|
---|
161 | : base(original, cloner) {
|
---|
162 | }
|
---|
163 |
|
---|
164 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
165 | return new ParameterlessPopulationPyramid(this, cloner);
|
---|
166 | }
|
---|
167 |
|
---|
168 | public ParameterlessPopulationPyramid() {
|
---|
169 | Parameters.Add(new FixedValueParameter<IntValue>(MaximumIterationsParameterName, "", new IntValue(Int32.MaxValue)));
|
---|
170 | Parameters.Add(new FixedValueParameter<IntValue>(MaximumEvaluationsParameterName, "", new IntValue(Int32.MaxValue)));
|
---|
171 | Parameters.Add(new FixedValueParameter<IntValue>(MaximumRuntimeParameterName, "The maximum runtime in seconds after which the algorithm stops. Use -1 to specify no limit for the runtime", new IntValue(3600)));
|
---|
172 | Parameters.Add(new FixedValueParameter<IntValue>(SeedParameterName, "The random seed used to initialize the new pseudo random number generator.", new IntValue(0)));
|
---|
173 | Parameters.Add(new FixedValueParameter<BoolValue>(SetSeedRandomlyParameterName, "True if the random seed should be set to a random value, otherwise false.", new BoolValue(true)));
|
---|
174 | }
|
---|
175 |
|
---|
176 | protected override void OnExecutionTimeChanged() {
|
---|
177 | base.OnExecutionTimeChanged();
|
---|
178 | if (CancellationTokenSource == null) return;
|
---|
179 | if (MaximumRuntime == -1) return;
|
---|
180 | if (ExecutionTime.TotalSeconds > MaximumRuntime) CancellationTokenSource.Cancel();
|
---|
181 | }
|
---|
182 |
|
---|
183 | private void AddIfUnique(BinaryVector solution, int level) {
|
---|
184 | // Don't add things you have seen
|
---|
185 | if (seen.Contains(solution)) return;
|
---|
186 | if (level == pyramid.Count) {
|
---|
187 | pyramid.Add(new Population(tracker.Length, random));
|
---|
188 | }
|
---|
189 | var copied = (BinaryVector)solution.Clone();
|
---|
190 | pyramid[level].Add(copied);
|
---|
191 | seen.Add(copied);
|
---|
192 | }
|
---|
193 |
|
---|
194 | // In the GECCO paper, Figure 1
|
---|
195 | private double iterate() {
|
---|
196 | // Create a random solution
|
---|
197 | BinaryVector solution = new BinaryVector(tracker.Length);
|
---|
198 | for (int i = 0; i < solution.Length; i++) {
|
---|
199 | solution[i] = random.Next(2) == 1;
|
---|
200 | }
|
---|
201 | double fitness = tracker.Evaluate(solution, random);
|
---|
202 | fitness = HillClimber.ImproveToLocalOptimum(tracker, solution, fitness, random);
|
---|
203 | AddIfUnique(solution, 0);
|
---|
204 |
|
---|
205 | for (int level = 0; level < pyramid.Count; level++) {
|
---|
206 | var current = pyramid[level];
|
---|
207 | double newFitness = LinkageCrossover.ImproveUsingTree(current.Tree, current.Solutions, solution, fitness, tracker, random);
|
---|
208 | // add it to the next level if its a strict fitness improvement
|
---|
209 | if (tracker.IsBetter(newFitness, fitness)) {
|
---|
210 | fitness = newFitness;
|
---|
211 | AddIfUnique(solution, level + 1);
|
---|
212 | }
|
---|
213 | }
|
---|
214 | return fitness;
|
---|
215 | }
|
---|
216 |
|
---|
217 | protected override void Initialize(CancellationToken cancellationToken) {
|
---|
218 | // Set up the algorithm
|
---|
219 | if (SetSeedRandomly) Seed = new System.Random().Next();
|
---|
220 | pyramid = new List<Population>();
|
---|
221 | seen.Clear();
|
---|
222 | random.Reset(Seed);
|
---|
223 | tracker = new EvaluationTracker(Problem, MaximumEvaluations);
|
---|
224 |
|
---|
225 | // Set up the results display
|
---|
226 | Results.Add(new Result("Iterations", new IntValue(0)));
|
---|
227 | Results.Add(new Result("Evaluations", new IntValue(0)));
|
---|
228 | Results.Add(new Result("Best Solution", new BinaryVector(tracker.BestSolution)));
|
---|
229 | Results.Add(new Result("Best Quality", new DoubleValue(tracker.BestQuality)));
|
---|
230 | Results.Add(new Result("Evaluation Best Solution Was Found", new IntValue(tracker.BestFoundOnEvaluation)));
|
---|
231 | var table = new DataTable("Qualities");
|
---|
232 | table.Rows.Add(new DataRow("Best Quality"));
|
---|
233 | var iterationRows = new DataRow("Iteration Quality");
|
---|
234 | iterationRows.VisualProperties.LineStyle = DataRowVisualProperties.DataRowLineStyle.Dot;
|
---|
235 | table.Rows.Add(iterationRows);
|
---|
236 | Results.Add(new Result("Qualities", table));
|
---|
237 |
|
---|
238 | table = new DataTable("Pyramid Levels");
|
---|
239 | table.Rows.Add(new DataRow("Levels"));
|
---|
240 | Results.Add(new Result("Pyramid Levels", table));
|
---|
241 |
|
---|
242 | table = new DataTable("Stored Solutions");
|
---|
243 | table.Rows.Add(new DataRow("Solutions"));
|
---|
244 | Results.Add(new Result("Stored Solutions", table));
|
---|
245 |
|
---|
246 | base.Initialize(cancellationToken);
|
---|
247 | }
|
---|
248 |
|
---|
249 | protected override void Run(CancellationToken cancellationToken) {
|
---|
250 | // Loop until iteration limit reached or canceled.
|
---|
251 | while (ResultsIterations < MaximumIterations) {
|
---|
252 | double fitness = double.NaN;
|
---|
253 |
|
---|
254 | try {
|
---|
255 | fitness = iterate();
|
---|
256 | ResultsIterations++;
|
---|
257 | cancellationToken.ThrowIfCancellationRequested();
|
---|
258 | } finally {
|
---|
259 | ResultsEvaluations = tracker.Evaluations;
|
---|
260 | ResultsBestSolution = new BinaryVector(tracker.BestSolution);
|
---|
261 | ResultsBestQuality = tracker.BestQuality;
|
---|
262 | ResultsBestFoundOnEvaluation = tracker.BestFoundOnEvaluation;
|
---|
263 | ResultsQualitiesBest.Values.Add(tracker.BestQuality);
|
---|
264 | ResultsQualitiesIteration.Values.Add(fitness);
|
---|
265 | ResultsLevels.Values.Add(pyramid.Count);
|
---|
266 | ResultsSolutions.Values.Add(seen.Count);
|
---|
267 | }
|
---|
268 | }
|
---|
269 | }
|
---|
270 | }
|
---|
271 | }
|
---|