Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/sources/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic/SymbolicRegressionProblemBase.cs @ 5439

Last change on this file since 5439 was 5386, checked in by mkommend, 14 years ago

Implemented SymbolicExpressionSymbolFrequencyAnalyzer (ticket #1398).
File size: 18.8 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2010 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 System.Linq;
25using HeuristicLab.Common;
26using HeuristicLab.Core;
27using HeuristicLab.Data;
28using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
29using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding.Analyzers;
30using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding.Creators;
31using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding.Interfaces;
32using HeuristicLab.Optimization;
33using HeuristicLab.Parameters;
34using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
35using HeuristicLab.PluginInfrastructure;
36using HeuristicLab.Problems.DataAnalysis.Regression.Symbolic.Analyzers;
37using HeuristicLab.Problems.DataAnalysis.Symbolic;
38
39namespace HeuristicLab.Problems.DataAnalysis.Regression.Symbolic {
40  [StorableClass]
41  public abstract class SymbolicRegressionProblemBase : DataAnalysisProblem, IProblem {
42
43    #region Parameter Properties
44    public new ValueParameter<SymbolicExpressionTreeCreator> SolutionCreatorParameter {
45      get { return (ValueParameter<SymbolicExpressionTreeCreator>)Parameters["SolutionCreator"]; }
46    }
47    IParameter IProblem.SolutionCreatorParameter {
48      get { return SolutionCreatorParameter; }
49    }
50    public ValueParameter<DoubleValue> LowerEstimationLimitParameter {
51      get { return (ValueParameter<DoubleValue>)Parameters["LowerEstimationLimit"]; }
52    }
53    public ValueParameter<DoubleValue> UpperEstimationLimitParameter {
54      get { return (ValueParameter<DoubleValue>)Parameters["UpperEstimationLimit"]; }
55    }
56    public ValueParameter<ISymbolicExpressionTreeInterpreter> SymbolicExpressionTreeInterpreterParameter {
57      get { return (ValueParameter<ISymbolicExpressionTreeInterpreter>)Parameters["SymbolicExpressionTreeInterpreter"]; }
58    }
59    public ValueParameter<ISymbolicExpressionGrammar> FunctionTreeGrammarParameter {
60      get { return (ValueParameter<ISymbolicExpressionGrammar>)Parameters["FunctionTreeGrammar"]; }
61    }
62    public ValueParameter<IntValue> MaxExpressionLengthParameter {
63      get { return (ValueParameter<IntValue>)Parameters["MaxExpressionLength"]; }
64    }
65    public ValueParameter<IntValue> MaxExpressionDepthParameter {
66      get { return (ValueParameter<IntValue>)Parameters["MaxExpressionDepth"]; }
67    }
68    public ValueParameter<IntValue> MaxFunctionDefiningBranchesParameter {
69      get { return (ValueParameter<IntValue>)Parameters["MaxFunctionDefiningBranches"]; }
70    }
71    public ValueParameter<IntValue> MaxFunctionArgumentsParameter {
72      get { return (ValueParameter<IntValue>)Parameters["MaxFunctionArguments"]; }
73    }
74    #endregion
75
76    #region Properties
77    public IntValue MaxExpressionLength {
78      get { return MaxExpressionLengthParameter.Value; }
79      set { MaxExpressionLengthParameter.Value = value; }
80    }
81    public IntValue MaxExpressionDepth {
82      get { return MaxExpressionDepthParameter.Value; }
83      set { MaxExpressionDepthParameter.Value = value; }
84    }
85    public IntValue MaxFunctionDefiningBranches {
86      get { return MaxFunctionDefiningBranchesParameter.Value; }
87      set { MaxFunctionDefiningBranchesParameter.Value = value; }
88    }
89    public IntValue MaxFunctionArguments {
90      get { return MaxFunctionArgumentsParameter.Value; }
91      set { MaxFunctionArgumentsParameter.Value = value; }
92    }
93    public new SymbolicExpressionTreeCreator SolutionCreator {
94      get { return SolutionCreatorParameter.Value; }
95      set { SolutionCreatorParameter.Value = value; }
96    }
97    ISolutionCreator IProblem.SolutionCreator {
98      get { return SolutionCreatorParameter.Value; }
99    }
100    public ISymbolicExpressionTreeInterpreter SymbolicExpressionTreeInterpreter {
101      get { return SymbolicExpressionTreeInterpreterParameter.Value; }
102      set { SymbolicExpressionTreeInterpreterParameter.Value = value; }
103    }
104    public DoubleValue LowerEstimationLimit {
105      get { return LowerEstimationLimitParameter.Value; }
106      set { LowerEstimationLimitParameter.Value = value; }
107    }
108    public DoubleValue UpperEstimationLimit {
109      get { return UpperEstimationLimitParameter.Value; }
110      set { UpperEstimationLimitParameter.Value = value; }
111    }
112
113    public ISymbolicExpressionGrammar FunctionTreeGrammar {
114      get { return (ISymbolicExpressionGrammar)FunctionTreeGrammarParameter.Value; }
115      private set { FunctionTreeGrammarParameter.Value = value; }
116    }
117    public override IEnumerable<IOperator> Operators {
118      get { return operators; }
119    }
120    public IEnumerable<ISymbolicRegressionAnalyzer> Analyzers {
121      get { return operators.OfType<ISymbolicRegressionAnalyzer>(); }
122    }
123    public DoubleValue PunishmentFactor {
124      get { return new DoubleValue(10.0); }
125    }
126    public IntValue TrainingSamplesStart {
127      get { return new IntValue(DataAnalysisProblemData.TrainingIndizes.First()); }
128    }
129    public IntValue TrainingSamplesEnd {
130      get {
131        int endIndex = (int)(DataAnalysisProblemData.TrainingIndizes.Count() * (1.0 - DataAnalysisProblemData.ValidationPercentage.Value) - 1);
132        if (endIndex < 0) endIndex = 0;
133        return new IntValue(DataAnalysisProblemData.TrainingIndizes.ElementAt(endIndex));
134      }
135    }
136    public IntValue ValidationSamplesStart {
137      get { return TrainingSamplesEnd; }
138    }
139    public IntValue ValidationSamplesEnd {
140      get { return new IntValue(DataAnalysisProblemData.TrainingIndizes.Last() + 1); }
141    }
142    public IntValue TestSamplesStart {
143      get { return DataAnalysisProblemData.TestSamplesStart; }
144    }
145    public IntValue TestSamplesEnd {
146      get { return DataAnalysisProblemData.TestSamplesEnd; }
147    }
148    #endregion
149
150    [Storable]
151    private List<IOperator> operators;
152
153    [StorableConstructor]
154    protected SymbolicRegressionProblemBase(bool deserializing) : base(deserializing) { }
155    protected SymbolicRegressionProblemBase(SymbolicRegressionProblemBase original, Cloner cloner)
156      : base(original, cloner) {
157      operators = original.operators.Select(x => (IOperator)cloner.Clone(x)).ToList();
158      RegisterParameterValueEvents();
159      RegisterParameterEvents();
160    }
161    public SymbolicRegressionProblemBase()
162      : base() {
163      SymbolicExpressionTreeCreator creator = new ProbabilisticTreeCreator();
164      var grammar = new FullFunctionalExpressionGrammar();
165      var globalGrammar = new GlobalSymbolicExpressionGrammar(grammar);
166      var interpreter = new SimpleArithmeticExpressionInterpreter();
167      Parameters.Add(new ValueParameter<SymbolicExpressionTreeCreator>("SolutionCreator", "The operator which should be used to create new symbolic regression solutions.", creator));
168      Parameters.Add(new ValueParameter<ISymbolicExpressionTreeInterpreter>("SymbolicExpressionTreeInterpreter", "The interpreter that should be used to evaluate the symbolic expression tree.", interpreter));
169      Parameters.Add(new ValueParameter<DoubleValue>("LowerEstimationLimit", "The lower limit for the estimated value that can be returned by the symbolic regression model.", new DoubleValue(double.NegativeInfinity)));
170      Parameters.Add(new ValueParameter<DoubleValue>("UpperEstimationLimit", "The upper limit for the estimated value that can be returned by the symbolic regression model.", new DoubleValue(double.PositiveInfinity)));
171      Parameters.Add(new ValueParameter<ISymbolicExpressionGrammar>("FunctionTreeGrammar", "The grammar that should be used for symbolic regression models.", globalGrammar));
172      Parameters.Add(new ValueParameter<IntValue>("MaxExpressionLength", "Maximal length of the symbolic expression.", new IntValue(100)));
173      Parameters.Add(new ValueParameter<IntValue>("MaxExpressionDepth", "Maximal depth of the symbolic expression. The minimum depth needed for the algorithm is 3 because two levels are reserved for the ProgramRoot and the Start symbol.", new IntValue(10)));
174      Parameters.Add(new ValueParameter<IntValue>("MaxFunctionDefiningBranches", "Maximal number of automatically defined functions.", new IntValue(0)));
175      Parameters.Add(new ValueParameter<IntValue>("MaxFunctionArguments", "Maximal number of arguments of automatically defined functions.", new IntValue(0)));
176
177      ParameterizeSolutionCreator();
178
179      UpdateGrammar();
180      UpdateEstimationLimits();
181      InitializeOperators();
182      RegisterParameterValueEvents();
183      RegisterParameterEvents();
184    }
185
186    private void RegisterParameterEvents() {
187      MaxFunctionArgumentsParameter.ValueChanged += new EventHandler(ArchitectureParameter_ValueChanged);
188      MaxFunctionDefiningBranchesParameter.ValueChanged += new EventHandler(ArchitectureParameter_ValueChanged);
189      MaxExpressionDepthParameter.ValueChanged += new EventHandler(MaxExpressionDepthParameter_ValueChanged);
190      SolutionCreatorParameter.ValueChanged += new EventHandler(SolutionCreatorParameter_ValueChanged);
191      FunctionTreeGrammarParameter.ValueChanged += new EventHandler(FunctionTreeGrammarParameter_ValueChanged);
192      SolutionCreator.SymbolicExpressionTreeParameter.ActualNameChanged += new EventHandler(SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged);
193    }
194
195    private void RegisterParameterValueEvents() {
196      MaxFunctionArgumentsParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
197      MaxFunctionDefiningBranchesParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
198      MaxExpressionDepthParameter.Value.ValueChanged += new EventHandler(MaxExpressionDepthParameterValue_ValueChanged);
199    }
200
201    #region event handling
202    protected override void OnDataAnalysisProblemChanged(EventArgs e) {
203      base.OnDataAnalysisProblemChanged(e);
204      // paritions could be changed
205      ParameterizeAnalyzers();
206      // input variables could have been changed
207      UpdateGrammar();
208      // estimation limits have to be recalculated
209      UpdateEstimationLimits();
210    }
211    protected virtual void OnArchitectureParameterChanged(EventArgs e) {
212      UpdateGrammar();
213    }
214    protected virtual void OnGrammarChanged() { UpdateGrammar(); }
215    protected virtual void OnOperatorsChanged(EventArgs e) { RaiseOperatorsChanged(e); }
216    protected virtual void OnSolutionCreatorChanged(EventArgs e) {
217      SolutionCreator.SymbolicExpressionTreeParameter.ActualNameChanged += new EventHandler(SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged);
218      ParameterizeSolutionCreator();
219      OnSolutionParameterNameChanged(e);
220      RaiseSolutionCreatorChanged(e);
221    }
222
223    protected virtual void OnSolutionParameterNameChanged(EventArgs e) {
224      ParameterizeAnalyzers();
225      ParameterizeOperators();
226    }
227
228    protected virtual void OnEvaluatorChanged(EventArgs e) {
229      RaiseEvaluatorChanged(e);
230    }
231    #endregion
232
233    #region event handlers
234    private void FunctionTreeGrammarParameter_ValueChanged(object sender, EventArgs e) {
235      if (!(FunctionTreeGrammar is GlobalSymbolicExpressionGrammar))
236        FunctionTreeGrammar = new GlobalSymbolicExpressionGrammar(FunctionTreeGrammar);
237      OnGrammarChanged();
238    }
239
240    private void SolutionCreatorParameter_ValueChanged(object sender, EventArgs e) {
241      OnSolutionCreatorChanged(e);
242    }
243    private void SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged(object sender, EventArgs e) {
244      OnSolutionParameterNameChanged(e);
245    }
246    private void ArchitectureParameter_ValueChanged(object sender, EventArgs e) {
247      MaxFunctionArgumentsParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
248      MaxFunctionDefiningBranchesParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
249      OnArchitectureParameterChanged(e);
250    }
251    private void ArchitectureParameterValue_ValueChanged(object sender, EventArgs e) {
252      OnArchitectureParameterChanged(e);
253    }
254
255    private void MaxExpressionDepthParameter_ValueChanged(object sender, EventArgs e) {
256      MaxExpressionDepthParameterValue_ValueChanged(sender, e);
257      MaxExpressionDepthParameter.Value.ValueChanged += MaxExpressionDepthParameterValue_ValueChanged;
258    }
259    private void MaxExpressionDepthParameterValue_ValueChanged(object sender, EventArgs e) {
260      if (MaxExpressionDepth != null && MaxExpressionDepth.Value < 3)
261        MaxExpressionDepth.Value = 3;
262    }
263    #endregion
264
265    #region Helpers
266    [StorableHook(HookType.AfterDeserialization)]
267    private void AfterDeserializationHook() {
268      // BackwardsCompatibility3.3
269      #region Backwards compatible code (remove with 3.4)
270      if (operators == null) InitializeOperators();
271      #endregion
272      RegisterParameterValueEvents();
273      RegisterParameterEvents();
274    }
275
276    protected void AddOperator(IOperator op) {
277      operators.Add(op);
278    }
279
280    private void UpdateGrammar() {
281      foreach (var varSymbol in FunctionTreeGrammar.Symbols.OfType<HeuristicLab.Problems.DataAnalysis.Symbolic.Symbols.Variable>()) {
282        varSymbol.VariableNames = DataAnalysisProblemData.InputVariables.CheckedItems.Select(x => x.Value.Value);
283      }
284      var globalGrammar = FunctionTreeGrammar as GlobalSymbolicExpressionGrammar;
285      if (globalGrammar != null) {
286        globalGrammar.MaxFunctionArguments = MaxFunctionArguments.Value;
287        globalGrammar.MaxFunctionDefinitions = MaxFunctionDefiningBranches.Value;
288      }
289    }
290
291    private void UpdateEstimationLimits() {
292      if (TrainingSamplesStart.Value < TrainingSamplesEnd.Value &&
293        DataAnalysisProblemData.Dataset.VariableNames.Contains(DataAnalysisProblemData.TargetVariable.Value)) {
294        var targetValues = DataAnalysisProblemData.Dataset.GetVariableValues(DataAnalysisProblemData.TargetVariable.Value, TrainingSamplesStart.Value, TrainingSamplesEnd.Value);
295        var mean = targetValues.Average();
296        var range = targetValues.Max() - targetValues.Min();
297        UpperEstimationLimit = new DoubleValue(mean + PunishmentFactor.Value * range);
298        LowerEstimationLimit = new DoubleValue(mean - PunishmentFactor.Value * range);
299      }
300    }
301
302    private void InitializeOperators() {
303      operators = new List<IOperator>();
304      operators.AddRange(ApplicationManager.Manager.GetInstances<ISymbolicExpressionTreeOperator>().OfType<IOperator>());
305      operators.Add(new SymbolicRegressionVariableFrequencyAnalyzer());
306      operators.Add(new SymbolicExpressionSymbolFrequencyAnalyzer());
307      operators.Add(new MinAverageMaxSymbolicExpressionTreeSizeAnalyzer());
308      operators.Add(new SymbolicRegressionModelQualityAnalyzer());
309      ParameterizeOperators();
310      ParameterizeAnalyzers();
311    }
312
313    private void ParameterizeSolutionCreator() {
314      SolutionCreator.SymbolicExpressionGrammarParameter.ActualName = FunctionTreeGrammarParameter.Name;
315      SolutionCreator.MaxTreeHeightParameter.ActualName = MaxExpressionDepthParameter.Name;
316      SolutionCreator.MaxTreeSizeParameter.ActualName = MaxExpressionLengthParameter.Name;
317      SolutionCreator.MaxFunctionArgumentsParameter.ActualName = MaxFunctionArgumentsParameter.Name;
318      SolutionCreator.MaxFunctionDefinitionsParameter.ActualName = MaxFunctionDefiningBranchesParameter.Name;
319    }
320
321    private void ParameterizeAnalyzers() {
322      foreach (var analyzer in Analyzers) {
323        analyzer.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
324        var symbolicRegressionModelQualityAnalyzer = analyzer as SymbolicRegressionModelQualityAnalyzer;
325        if (symbolicRegressionModelQualityAnalyzer != null) {
326          symbolicRegressionModelQualityAnalyzer.ProblemDataParameter.ActualName = DataAnalysisProblemDataParameter.Name;
327          symbolicRegressionModelQualityAnalyzer.UpperEstimationLimitParameter.ActualName = UpperEstimationLimitParameter.Name;
328          symbolicRegressionModelQualityAnalyzer.LowerEstimationLimitParameter.ActualName = LowerEstimationLimitParameter.Name;
329          symbolicRegressionModelQualityAnalyzer.SymbolicExpressionTreeInterpreterParameter.ActualName = SymbolicExpressionTreeInterpreterParameter.Name;
330          symbolicRegressionModelQualityAnalyzer.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
331        }
332        var varFreqAnalyzer = analyzer as SymbolicRegressionVariableFrequencyAnalyzer;
333        if (varFreqAnalyzer != null) {
334          varFreqAnalyzer.ProblemDataParameter.ActualName = DataAnalysisProblemDataParameter.Name;
335        }
336      }
337      foreach (ISymbolicExpressionTreeAnalyzer analyzer in Operators.OfType<ISymbolicExpressionTreeAnalyzer>()) {
338        analyzer.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
339      }
340    }
341
342    private void ParameterizeOperators() {
343      foreach (ISymbolicExpressionTreeOperator op in Operators.OfType<ISymbolicExpressionTreeOperator>()) {
344        op.MaxTreeHeightParameter.ActualName = MaxExpressionDepthParameter.Name;
345        op.MaxTreeSizeParameter.ActualName = MaxExpressionLengthParameter.Name;
346        op.SymbolicExpressionGrammarParameter.ActualName = FunctionTreeGrammarParameter.Name;
347      }
348      foreach (ISymbolicExpressionTreeCrossover op in Operators.OfType<ISymbolicExpressionTreeCrossover>()) {
349        op.ParentsParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
350        op.ChildParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
351      }
352      foreach (ISymbolicExpressionTreeManipulator op in Operators.OfType<ISymbolicExpressionTreeManipulator>()) {
353        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
354      }
355      foreach (ISymbolicExpressionTreeArchitectureManipulator op in Operators.OfType<ISymbolicExpressionTreeArchitectureManipulator>()) {
356        op.MaxFunctionArgumentsParameter.ActualName = MaxFunctionArgumentsParameter.Name;
357        op.MaxFunctionDefinitionsParameter.ActualName = MaxFunctionDefiningBranchesParameter.Name;
358      }
359    }
360    #endregion
361  }
362}
Note: See TracBrowser for help on using the repository browser.