1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2013 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
4 | *
|
---|
5 | * This file is part of HeuristicLab.
|
---|
6 | *
|
---|
7 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
8 | * it under the terms of the GNU General Public License as published by
|
---|
9 | * the Free Software Foundation, either version 3 of the License, or
|
---|
10 | * (at your option) any later version.
|
---|
11 | *
|
---|
12 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
15 | * GNU General Public License for more details.
|
---|
16 | *
|
---|
17 | * You should have received a copy of the GNU General Public License
|
---|
18 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
19 | */
|
---|
20 | #endregion
|
---|
21 |
|
---|
22 | using HeuristicLab.Common;
|
---|
23 | using HeuristicLab.Core;
|
---|
24 | using HeuristicLab.Data;
|
---|
25 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
26 | using HeuristicLab.Problems.ArtificialAnt;
|
---|
27 | using HeuristicLab.Operators;
|
---|
28 | using HeuristicLab.Optimization;
|
---|
29 | using HeuristicLab.Parameters;
|
---|
30 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
31 | using HeuristicLab.Encodings.IntegerVectorEncoding;
|
---|
32 | using System.Collections.Generic;
|
---|
33 | using System.Linq;
|
---|
34 | using HeuristicLab.Random;
|
---|
35 |
|
---|
36 | namespace HeuristicLab.Problems.GrammaticalEvolution {
|
---|
37 | [Item("GEArtificialAntEvaluator", "Evaluates an artificial ant solution, implemented in Grammatical Evolution.")]
|
---|
38 | [StorableClass]
|
---|
39 | public class GEEvaluator : SingleSuccessorOperator,
|
---|
40 | ISingleObjectiveEvaluator, ISymbolicExpressionTreeGrammarBasedOperator {
|
---|
41 |
|
---|
42 | #region Parameter Properties
|
---|
43 | public ILookupParameter<DoubleValue> QualityParameter {
|
---|
44 | get { return (ILookupParameter<DoubleValue>)Parameters["Quality"]; }
|
---|
45 | }
|
---|
46 | public ILookupParameter<IntegerVector> IntegerVectorParameter {
|
---|
47 | get { return (ILookupParameter<IntegerVector>)Parameters["IntegerVector"]; }
|
---|
48 | }
|
---|
49 | public ILookupParameter<SymbolicExpressionTree> SymbolicExpressionTreeParameter {
|
---|
50 | get { return (ILookupParameter<SymbolicExpressionTree>)Parameters["SymbolicExpressionTree"]; }
|
---|
51 | }
|
---|
52 | public ILookupParameter<BoolMatrix> WorldParameter {
|
---|
53 | get { return (ILookupParameter<BoolMatrix>)Parameters["World"]; }
|
---|
54 | }
|
---|
55 | public ILookupParameter<IntValue> MaxTimeStepsParameter {
|
---|
56 | get { return (ILookupParameter<IntValue>)Parameters["MaxTimeSteps"]; }
|
---|
57 | }
|
---|
58 | public IValueLookupParameter<ISymbolicExpressionGrammar> SymbolicExpressionTreeGrammarParameter {
|
---|
59 | get { return (IValueLookupParameter<ISymbolicExpressionGrammar>)Parameters["SymbolicExpressionTreeGrammar"]; }
|
---|
60 | }
|
---|
61 | #endregion
|
---|
62 |
|
---|
63 | [StorableConstructor]
|
---|
64 | protected GEEvaluator(bool deserializing) : base(deserializing) { }
|
---|
65 | protected GEEvaluator(GEEvaluator original, Cloner cloner) : base(original, cloner) { }
|
---|
66 | public override IDeepCloneable Clone(Cloner cloner) { return new GEEvaluator(this, cloner); }
|
---|
67 | public GEEvaluator()
|
---|
68 | : base() {
|
---|
69 | Parameters.Add(new LookupParameter<DoubleValue>("Quality", "The quality of the evaluated artificial ant solution."));
|
---|
70 | Parameters.Add(new LookupParameter<IntegerVector>("IntegerVector", "The artificial ant solution encoded as an integer vector genome."));
|
---|
71 | Parameters.Add(new LookupParameter<SymbolicExpressionTree>("SymbolicExpressionTree", "The artificial ant solution encoded as a symbolic expression tree that should be evaluated"));
|
---|
72 | Parameters.Add(new LookupParameter<BoolMatrix>("World", "The world for the artificial ant with scattered food items."));
|
---|
73 | Parameters.Add(new LookupParameter<IntValue>("MaxTimeSteps", "The maximal number of time steps that the artificial ant should be simulated."));
|
---|
74 | Parameters.Add(new ValueLookupParameter<ISymbolicExpressionGrammar>("SymbolicExpressionTreeGrammar", "The tree grammar that defines the correct syntax of symbolic expression trees that should be created."));
|
---|
75 | }
|
---|
76 |
|
---|
77 | public sealed override IOperation Apply() {
|
---|
78 | SymbolicExpressionTree expression = MapIntegerVectorToSymbolicExpressionTree();
|
---|
79 | BoolMatrix world = WorldParameter.ActualValue;
|
---|
80 | IntValue maxTimeSteps = MaxTimeStepsParameter.ActualValue;
|
---|
81 |
|
---|
82 | AntInterpreter interpreter = new AntInterpreter();
|
---|
83 | interpreter.MaxTimeSteps = maxTimeSteps.Value;
|
---|
84 | interpreter.World = world;
|
---|
85 | interpreter.Expression = expression;
|
---|
86 | interpreter.Run();
|
---|
87 |
|
---|
88 | QualityParameter.ActualValue = new DoubleValue(interpreter.FoodEaten);
|
---|
89 | return null;
|
---|
90 | }
|
---|
91 |
|
---|
92 |
|
---|
93 | /// <summary>
|
---|
94 | /// Maps an integer vector to a symbolic expression tree, using a
|
---|
95 | /// genotype-to-phenotype mapper.
|
---|
96 | /// </summary>
|
---|
97 | /// <returns>solution tree</returns>
|
---|
98 | private SymbolicExpressionTree MapIntegerVectorToSymbolicExpressionTree() {
|
---|
99 |
|
---|
100 | ISymbolicExpressionGrammar grammar = SymbolicExpressionTreeGrammarParameter.ActualValue;
|
---|
101 | SymbolicExpressionTree tree = new SymbolicExpressionTree();
|
---|
102 | IntegerVector integerVectorGenome = IntegerVectorParameter.ActualValue;
|
---|
103 | var rootNode = (SymbolicExpressionTreeTopLevelNode)grammar.ProgramRootSymbol.CreateTreeNode();
|
---|
104 | var startNode = (SymbolicExpressionTreeTopLevelNode)grammar.StartSymbol.CreateTreeNode();
|
---|
105 | rootNode.AddSubtree(startNode);
|
---|
106 | tree.Root = rootNode;
|
---|
107 |
|
---|
108 | int genotypeIndex = 0;
|
---|
109 | int currSubtreeCount = 1;
|
---|
110 |
|
---|
111 | MapGenoToPhenoDepthFirstRec(startNode, integerVectorGenome,
|
---|
112 | grammar, integerVectorGenome.Length,
|
---|
113 | ref genotypeIndex, ref currSubtreeCount);
|
---|
114 |
|
---|
115 | SymbolicExpressionTreeParameter.ActualValue = tree;
|
---|
116 | return tree;
|
---|
117 | }
|
---|
118 |
|
---|
119 |
|
---|
120 | /// <summary>
|
---|
121 | /// Genotype-to-Phenotype mapper (recursive depth-first approach).
|
---|
122 | /// Appends maximum allowed children (non-terminal symbols) to
|
---|
123 | /// <paramref name="currentNode"/>, as long as <paramref name="currSubtreeCount"/>
|
---|
124 | /// doesn't exceed <paramref name="maxSubtreeCount"/>.
|
---|
125 | /// If at most <paramref name="maxSubtreeCount"/> subtrees were created,
|
---|
126 | /// each non-full node is filled with randomly chosen nodes
|
---|
127 | /// (non-terminal and terminal), and each non-terminal node is again filled with a terminal node.
|
---|
128 | /// </summary>
|
---|
129 | /// <param name="currentNode">current parent node</param>
|
---|
130 | /// <param name="integerVectorGenome">integer vector used for mapping</param>
|
---|
131 | /// <param name="grammar">grammar definition to determine the allowed child symbols for currentNode </param>
|
---|
132 | /// <param name="maxSubtreeCount">maximum allowed subtrees (= number of used genomes)</param>
|
---|
133 | /// <param name="genotypeIndex">current index in integer vector</param>
|
---|
134 | /// <param name="currSubtreeCount">number of already determined subtrees (filled or still incomplete)</param>
|
---|
135 | private void MapGenoToPhenoDepthFirstRec(ISymbolicExpressionTreeNode currentNode,
|
---|
136 | IntegerVector integerVectorGenome,
|
---|
137 | ISymbolicExpressionGrammar grammar,
|
---|
138 | int maxSubtreeCount,
|
---|
139 | ref int genotypeIndex,
|
---|
140 | ref int currSubtreeCount) {
|
---|
141 | if (currSubtreeCount < maxSubtreeCount) {
|
---|
142 |
|
---|
143 | var newNode = GetNewChildNode(currentNode, integerVectorGenome, grammar, genotypeIndex);
|
---|
144 |
|
---|
145 | if ((currSubtreeCount + newNode.Symbol.MaximumArity) > maxSubtreeCount) {
|
---|
146 | // TODO: maybe check, if there is any node, which fits in the tree yet
|
---|
147 | currentNode.AddSubtree(GetRandomTerminalNode(currentNode, grammar));
|
---|
148 | } else {
|
---|
149 | currentNode.AddSubtree(newNode);
|
---|
150 | genotypeIndex++;
|
---|
151 | currSubtreeCount += newNode.Symbol.MaximumArity;
|
---|
152 |
|
---|
153 | while (newNode.Symbol.MaximumArity > newNode.SubtreeCount) {
|
---|
154 | MapGenoToPhenoDepthFirstRec(newNode, integerVectorGenome,
|
---|
155 | grammar, maxSubtreeCount,
|
---|
156 | ref genotypeIndex, ref currSubtreeCount);
|
---|
157 | }
|
---|
158 | }
|
---|
159 |
|
---|
160 | } else {
|
---|
161 | while (currentNode.Symbol.MaximumArity > currentNode.SubtreeCount) {
|
---|
162 | var newNode = GetNewChildNode(currentNode, integerVectorGenome, grammar, genotypeIndex);
|
---|
163 | currentNode.AddSubtree(newNode);
|
---|
164 | genotypeIndex++;
|
---|
165 | while (newNode.Symbol.MaximumArity > newNode.SubtreeCount) {
|
---|
166 | newNode.AddSubtree(GetRandomTerminalNode(newNode, grammar));
|
---|
167 | }
|
---|
168 | }
|
---|
169 | }
|
---|
170 | }
|
---|
171 |
|
---|
172 |
|
---|
173 | /// <summary>
|
---|
174 | /// Randomly returns a terminal node for the given <paramref name="parentNode"/>.
|
---|
175 | /// (A terminal has got a minimum and maximum arity of 0.)
|
---|
176 | /// </summary>
|
---|
177 | /// <param name="parentNode">parent node for which a child node is returned randomly</param>
|
---|
178 | /// <param name="grammar">grammar definition to determine the allowed child symbols for parentNode</param>
|
---|
179 | /// <returns>randomly chosen terminal node with arity 0</returns>
|
---|
180 | private ISymbolicExpressionTreeNode GetRandomTerminalNode(ISymbolicExpressionTreeNode parentNode,
|
---|
181 | ISymbolicExpressionGrammar grammar) {
|
---|
182 | var possibleSymbolsList = from s in grammar.GetAllowedChildSymbols(parentNode.Symbol)
|
---|
183 | where s.MaximumArity == 0
|
---|
184 | where s.MinimumArity == 0
|
---|
185 | select s;
|
---|
186 | // TODO: Check, if symbol list is empty (no terminal nodes found) - what should happen?
|
---|
187 | return possibleSymbolsList.SelectRandom(new MersenneTwister()).CreateTreeNode();
|
---|
188 | }
|
---|
189 |
|
---|
190 |
|
---|
191 | /// <summary>
|
---|
192 | /// Utility method, which returns the number of elements of <paramref name="symbolList"/>.
|
---|
193 | /// </summary>
|
---|
194 | /// <param name="symbolList">enumerable symbol list to count the elements for</param>
|
---|
195 | /// <returns>number of elements in parameter symbolList</returns>
|
---|
196 | private int GetNumberOfAllowedChildSymbols(IEnumerable<ISymbol> symbolList) {
|
---|
197 | int count = 0;
|
---|
198 | using (IEnumerator<ISymbol> enumerator = symbolList.GetEnumerator()) {
|
---|
199 | while (enumerator.MoveNext()) {
|
---|
200 | count++;
|
---|
201 | }
|
---|
202 | }
|
---|
203 | return count;
|
---|
204 | }
|
---|
205 |
|
---|
206 |
|
---|
207 | /// <summary>
|
---|
208 | /// Returns a randomly chosen child node for the given <paramref name="parentNode"/>.
|
---|
209 | /// </summary>
|
---|
210 | /// <param name="parentNode">parent node to find a child node randomly for</param>
|
---|
211 | /// <param name="integerVectorGenome">integer vector to map to production rules</param>
|
---|
212 | /// <param name="grammar">grammar definition used to define the allowed child symbols</param>
|
---|
213 | /// <param name="genotypeIndex">index in the integer vector; can be greater than vector length</param>
|
---|
214 | /// <returns></returns>
|
---|
215 | private ISymbolicExpressionTreeNode GetNewChildNode(ISymbolicExpressionTreeNode parentNode,
|
---|
216 | IntegerVector integerVectorGenome,
|
---|
217 | ISymbolicExpressionGrammar grammar,
|
---|
218 | int genotypeIndex) {
|
---|
219 |
|
---|
220 | var symbolList = grammar.GetAllowedChildSymbols(parentNode.Symbol);
|
---|
221 | int prodRuleCount = GetNumberOfAllowedChildSymbols(symbolList);
|
---|
222 | int prodRuleIndex = integerVectorGenome[genotypeIndex % integerVectorGenome.Length] % prodRuleCount;
|
---|
223 | int currentIndex = 0;
|
---|
224 |
|
---|
225 | using (IEnumerator<ISymbol> enumerator = symbolList.GetEnumerator()) {
|
---|
226 | while (enumerator.MoveNext() && (currentIndex != prodRuleIndex)) {
|
---|
227 | currentIndex++;
|
---|
228 | }
|
---|
229 | return enumerator.Current.CreateTreeNode();
|
---|
230 | }
|
---|
231 | }
|
---|
232 | }
|
---|
233 | } |
---|