Free cookie consent management tool by TermsFeed Policy Generator

source: branches/GrammaticalEvolution/HeuristicLab.Problems.GrammaticalEvolution/Mappers/DepthFirstMapper.cs @ 10228

Last change on this file since 10228 was 10228, checked in by sawinkle, 10 years ago

#2109: Updated DepthFirstMapper and abstract base class GenotypeToPhenotypeMapper:

  • Added new method SampleArity() to GenotypeToPhenotypeMapper to determine a random arity for a given node, depending on a maximum allowed arity.
  • Replaced the recursive depth-first mapping approach by a iterative one, which uses a stack of <node, arity> tuples. The recursive approach only generated trees with very small subtrees depending on the minimumArity of each node. Now, the iterative one uses the SampleArity() method and pushes/pops the <node, arity> tuples from/to the used stack. Therefore, it is not necessary to only allow the minimumArity, but also to deal with arbitrarily sampled arities per node.
File size: 8.9 KB
Line 
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
22using System;
23using System.Collections.Generic;
24using HeuristicLab.Common;
25using HeuristicLab.Core;
26using HeuristicLab.Encodings.IntegerVectorEncoding;
27using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
28using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
29using HeuristicLab.Random;
30
31namespace HeuristicLab.Problems.GrammaticalEvolution {
32  /// <summary>
33  /// DepthFirstMapper
34  /// </summary>
35  [Item("DepthFirstMapper", "")]
36  [StorableClass]
37  public class DepthFirstMapper : GenotypeToPhenotypeMapper {
38
39    [StorableConstructor]
40    protected DepthFirstMapper(bool deserializing) : base(deserializing) { }
41    protected DepthFirstMapper(DepthFirstMapper original, Cloner cloner) : base(original, cloner) { }
42    public DepthFirstMapper() : base() { }
43
44    public override IDeepCloneable Clone(Cloner cloner) {
45      return new DepthFirstMapper(this, cloner);
46    }
47
48
49    /// <summary>
50    /// Maps a genotype (an integer vector) to a phenotype (a symbolic expression tree).
51    /// Depth-first approach.
52    /// </summary>
53    /// <param name="grammar">grammar definition</param>
54    /// <param name="genotype">integer vector, which should be mapped to a tree</param>
55    /// <returns>phenotype (a symbolic expression tree)</returns>
56    public override SymbolicExpressionTree Map(ISymbolicExpressionGrammar grammar,
57                                               IntegerVector genotype) {
58
59      SymbolicExpressionTree tree = new SymbolicExpressionTree();
60      var rootNode = (SymbolicExpressionTreeTopLevelNode)grammar.ProgramRootSymbol.CreateTreeNode();
61      if (rootNode.HasLocalParameters) rootNode.ResetLocalParameters(new MersenneTwister());
62      var startNode = (SymbolicExpressionTreeTopLevelNode)grammar.StartSymbol.CreateTreeNode();
63      if (startNode.HasLocalParameters) startNode.ResetLocalParameters(new MersenneTwister());
64      rootNode.AddSubtree(startNode);
65      tree.Root = rootNode;
66
67      //int genotypeIndex = 0;
68      //int currSubtreeCount = 1;
69      //MapDepthFirstRecursively(startNode, genotype,
70      //                         grammar, genotype.Length,
71      //                         ref genotypeIndex, ref currSubtreeCount,
72      //                         new MersenneTwister());
73
74      MapDepthFirstIteratively(startNode, genotype, grammar,
75                               genotype.Length, new MersenneTwister());
76      return tree;
77    }
78
79
80    /// <summary>
81    /// Genotype-to-Phenotype mapper (recursive depth-first approach).
82    /// Appends maximum allowed children (non-terminal symbols) to
83    /// <paramref name="currentNode"/>, as long as <paramref name="currSubtreeCount"/>
84    /// doesn't exceed <paramref name="maxSubtreeCount"/>.
85    /// If at most <paramref name="maxSubtreeCount"/> subtrees were created,
86    /// each non-full node is filled with randomly chosen nodes
87    /// (non-terminal and terminal), and each non-terminal node is again filled with a terminal node.
88    /// </summary>
89    /// <param name="currentNode">current parent node</param>
90    /// <param name="genotype">integer vector, which should be mapped to a tree</param>
91    /// <param name="grammar">grammar to determine the allowed child symbols for currentNode </param>
92    /// <param name="maxSubtreeCount">maximum allowed subtrees (= number of used genomes)</param>
93    /// <param name="genotypeIndex">current index in integer vector</param>
94    /// <param name="currSubtreeCount">number of already determined subtrees (filled or still incomplete)</param>
95    private void MapDepthFirstRecursively(ISymbolicExpressionTreeNode currentNode,
96                                          IntegerVector genotype,
97                                          ISymbolicExpressionGrammar grammar,
98                                          int maxSubtreeCount,
99                                          ref int genotypeIndex,
100                                          ref int currSubtreeCount,
101                                          IRandom random) {
102
103      // TODO: check, if method calls of GetNewChildNode() and GetRandomTerminalNode() don't return null
104      if (currSubtreeCount < maxSubtreeCount) {
105
106        var newNode = GetNewChildNode(currentNode, genotype, grammar, genotypeIndex, random);
107
108        if ((currSubtreeCount + newNode.Symbol.MinimumArity) > maxSubtreeCount) {
109          // TODO: maybe check, if there is any node, which fits in the tree yet
110          currentNode.AddSubtree(GetRandomTerminalNode(currentNode, grammar, random));
111        } else {
112          currentNode.AddSubtree(newNode);
113          genotypeIndex++;
114          currSubtreeCount += newNode.Symbol.MinimumArity;
115
116          while (newNode.Symbol.MinimumArity > newNode.SubtreeCount) {
117            MapDepthFirstRecursively(newNode, genotype,
118                                     grammar, maxSubtreeCount,
119                                     ref genotypeIndex, ref currSubtreeCount, random);
120          }
121        }
122
123      } else {
124        while (currentNode.Symbol.MinimumArity > currentNode.SubtreeCount) {
125          var newNode = GetNewChildNode(currentNode, genotype, grammar, genotypeIndex, random);
126          currentNode.AddSubtree(newNode);
127          genotypeIndex++;
128          while (newNode.Symbol.MinimumArity > newNode.SubtreeCount) {
129            newNode.AddSubtree(GetRandomTerminalNode(newNode, grammar, random));
130          }
131        }
132      }
133    }
134
135
136    /// <summary>
137    /// Genotype-to-Phenotype mapper (iterative depth-first approach).
138    /// </summary>
139    /// <param name="startNode">first node of the tree with arity 1</param>
140    /// <param name="genotype">integer vector, which should be mapped to a tree</param>
141    /// <param name="grammar">grammar to determine the allowed child symbols for each node</param>
142    /// <param name="maxSubtreeCount">maximum allowed subtrees (= number of used genomes)</param>
143    /// <param name="random">random number generator</param>
144    private void MapDepthFirstIteratively(ISymbolicExpressionTreeNode startNode,
145                                          IntegerVector genotype,
146                                          ISymbolicExpressionGrammar grammar,
147                                          int maxSubtreeCount, IRandom random) {
148
149      Stack<Tuple<ISymbolicExpressionTreeNode, int>> stack
150        = new Stack<Tuple<ISymbolicExpressionTreeNode, int>>(); // tuples of <node, arity>
151
152      int genotypeIndex = 0;
153      int currSubtreeCount = 1;
154
155      stack.Push(new Tuple<ISymbolicExpressionTreeNode, int>(startNode, 1));
156
157      while ((currSubtreeCount < maxSubtreeCount) && (stack.Count > 0)) {
158
159        // get next node from stack and re-push it, if this node still has unhandled subtrees ...
160        Tuple<ISymbolicExpressionTreeNode, int> current = stack.Pop();
161        if (current.Item2 > 1) {
162          stack.Push(new Tuple<ISymbolicExpressionTreeNode, int>(current.Item1, current.Item2 - 1));
163        }
164
165        var newNode = GetNewChildNode(current.Item1, genotype, grammar, genotypeIndex, random);
166        int arity = SampleArity(random, newNode, maxSubtreeCount - currSubtreeCount);
167
168        if (arity < 0) {
169          current.Item1.AddSubtree(GetRandomTerminalNode(current.Item1, grammar, random));
170        } else {
171          current.Item1.AddSubtree(newNode);
172          genotypeIndex++;
173          currSubtreeCount += arity;
174          if (arity > 0) {
175            // new node has subtrees so push it onto the stack
176            stack.Push(new Tuple<ISymbolicExpressionTreeNode, int>(newNode, arity));
177          }
178        }
179      }
180
181      // maximum allowed subtree count was already reached, but there are still
182      // incomplete subtrees (non-terminal symbols) in the tree
183      // -> fill them with terminal symbols
184      while (stack.Count > 0) {
185        Tuple<ISymbolicExpressionTreeNode, int> current = stack.Pop();
186        if (current.Item2 > 1) {
187          stack.Push(new Tuple<ISymbolicExpressionTreeNode, int>(current.Item1, current.Item2 - 1));
188        }
189        current.Item1.AddSubtree(GetRandomTerminalNode(current.Item1, grammar, random));
190      }
191    }
192  }
193}
Note: See TracBrowser for help on using the repository browser.