Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SymbolicRegressionPruningOperator.cs @ 14349

Last change on this file since 14349 was 14185, checked in by swagner, 8 years ago

#2526: Updated year of copyrights in license headers

File size: 5.8 KB
Line 
1#region License Information
2
3/* HeuristicLab
4 * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
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
22#endregion
23
24using System.Collections.Generic;
25using System.Linq;
26using HeuristicLab.Common;
27using HeuristicLab.Core;
28using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
29using HeuristicLab.Parameters;
30using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
31
32namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression {
33  [StorableClass]
34  [Item("SymbolicRegressionPruningOperator", "An operator which prunes symbolic regression trees.")]
35  public class SymbolicRegressionPruningOperator : SymbolicDataAnalysisExpressionPruningOperator {
36    private const string EvaluatorParameterName = "Evaluator";
37
38    #region parameter properties
39    public ILookupParameter<ISymbolicRegressionSingleObjectiveEvaluator> EvaluatorParameter {
40      get { return (ILookupParameter<ISymbolicRegressionSingleObjectiveEvaluator>)Parameters[EvaluatorParameterName]; }
41    }
42    #endregion
43
44    protected SymbolicRegressionPruningOperator(SymbolicRegressionPruningOperator original, Cloner cloner)
45      : base(original, cloner) {
46    }
47    public override IDeepCloneable Clone(Cloner cloner) {
48      return new SymbolicRegressionPruningOperator(this, cloner);
49    }
50
51    [StorableConstructor]
52    protected SymbolicRegressionPruningOperator(bool deserializing) : base(deserializing) { }
53
54    public SymbolicRegressionPruningOperator(ISymbolicDataAnalysisSolutionImpactValuesCalculator impactValuesCalculator)
55      : base(impactValuesCalculator) {
56      Parameters.Add(new LookupParameter<ISymbolicRegressionSingleObjectiveEvaluator>(EvaluatorParameterName));
57    }
58
59    [StorableHook(HookType.AfterDeserialization)]
60    private void AfterDeserialization() {
61      // BackwardsCompatibility3.3
62      #region Backwards compatible code, remove with 3.4
63      base.ImpactValuesCalculator = new SymbolicRegressionSolutionImpactValuesCalculator();
64      if (!Parameters.ContainsKey(EvaluatorParameterName)) {
65        Parameters.Add(new LookupParameter<ISymbolicRegressionSingleObjectiveEvaluator>(EvaluatorParameterName));
66      }
67      #endregion
68    }
69
70    protected override ISymbolicDataAnalysisModel CreateModel(ISymbolicExpressionTree tree, ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, IDataAnalysisProblemData problemData, DoubleLimit estimationLimits) {
71      var regressionProblemData = (IRegressionProblemData)problemData;
72      return new SymbolicRegressionModel(regressionProblemData.TargetVariable, tree, interpreter, estimationLimits.Lower, estimationLimits.Upper);
73    }
74
75    protected override double Evaluate(IDataAnalysisModel model) {
76      var regressionModel = (ISymbolicRegressionModel)model;
77      var regressionProblemData = (IRegressionProblemData)ProblemDataParameter.ActualValue;
78      var evaluator = EvaluatorParameter.ActualValue;
79      var fitnessEvaluationPartition = FitnessCalculationPartitionParameter.ActualValue;
80      var rows = Enumerable.Range(fitnessEvaluationPartition.Start, fitnessEvaluationPartition.Size);
81      return evaluator.Evaluate(this.ExecutionContext, regressionModel.SymbolicExpressionTree, regressionProblemData, rows);
82    }
83
84    public static ISymbolicExpressionTree Prune(ISymbolicExpressionTree tree, SymbolicRegressionSolutionImpactValuesCalculator impactValuesCalculator, ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, IRegressionProblemData problemData, DoubleLimit estimationLimits, IEnumerable<int> rows, double nodeImpactThreshold = 0.0, bool pruneOnlyZeroImpactNodes = false) {
85      var clonedTree = (ISymbolicExpressionTree)tree.Clone();
86      var model = new SymbolicRegressionModel(problemData.TargetVariable, clonedTree, interpreter, estimationLimits.Lower, estimationLimits.Upper);
87      var nodes = clonedTree.Root.GetSubtree(0).GetSubtree(0).IterateNodesPrefix().ToList(); // skip the nodes corresponding to the ProgramRootSymbol and the StartSymbol
88
89      double qualityForImpactsCalculation = double.NaN; // pass a NaN value initially so the impact calculator will calculate the quality
90
91      for (int i = 0; i < nodes.Count; ++i) {
92        var node = nodes[i];
93        if (node is ConstantTreeNode) continue;
94
95        double impactValue, replacementValue;
96        double newQualityForImpactsCalculation;
97        impactValuesCalculator.CalculateImpactAndReplacementValues(model, node, problemData, rows, out impactValue, out replacementValue, out newQualityForImpactsCalculation, qualityForImpactsCalculation);
98
99        if (pruneOnlyZeroImpactNodes && !impactValue.IsAlmost(0.0)) continue;
100        if (!pruneOnlyZeroImpactNodes && impactValue > nodeImpactThreshold) continue;
101
102        var constantNode = (ConstantTreeNode)node.Grammar.GetSymbol("Constant").CreateTreeNode();
103        constantNode.Value = replacementValue;
104
105        ReplaceWithConstant(node, constantNode);
106        i += node.GetLength() - 1; // skip subtrees under the node that was folded
107
108        qualityForImpactsCalculation = newQualityForImpactsCalculation;
109      }
110      return model.SymbolicExpressionTree;
111    }
112  }
113}
Note: See TracBrowser for help on using the repository browser.