Free cookie consent management tool by TermsFeed Policy Generator

source: branches/DataAnalysis/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic/SymbolicRegressionProblemBase.cs @ 4275

Last change on this file since 4275 was 4225, checked in by gkronber, 14 years ago

Minor adaption of existing symbolic regression classes. #1142.
File size: 18.5 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    }
116    public override IEnumerable<IOperator> Operators {
117      get { return operators; }
118    }
119    public IEnumerable<ISymbolicRegressionAnalyzer> Analyzers {
120      get { return operators.OfType<ISymbolicRegressionAnalyzer>(); }
121    }
122    public DoubleValue PunishmentFactor {
123      get { return new DoubleValue(10.0); }
124    }
125    public IntValue TrainingSamplesStart {
126      get { return new IntValue(DataAnalysisProblemData.TrainingSamplesStart.Value); }
127    }
128    public IntValue TrainingSamplesEnd {
129      get {
130        return new IntValue((DataAnalysisProblemData.TrainingSamplesStart.Value +
131          DataAnalysisProblemData.TrainingSamplesEnd.Value) / 2);
132      }
133    }
134    public IntValue ValidationSamplesStart {
135      get { return TrainingSamplesEnd; }
136    }
137    public IntValue ValidationSamplesEnd {
138      get { return new IntValue(DataAnalysisProblemData.TrainingSamplesEnd.Value); }
139    }
140    public IntValue TestSamplesStart {
141      get { return DataAnalysisProblemData.TestSamplesStart; }
142    }
143    public IntValue TestSamplesEnd {
144      get { return DataAnalysisProblemData.TestSamplesEnd; }
145    }
146    #endregion
147
148    [Storable]
149    private List<IOperator> operators;
150
151    [StorableConstructor]
152    protected SymbolicRegressionProblemBase(bool deserializing) : base(deserializing) { }
153    public SymbolicRegressionProblemBase()
154      : base() {
155      SymbolicExpressionTreeCreator creator = new ProbabilisticTreeCreator();
156      var grammar = new FullFunctionalExpressionGrammar();
157      var globalGrammar = new GlobalSymbolicExpressionGrammar(grammar);
158      var interpreter = new SimpleArithmeticExpressionInterpreter();
159      Parameters.Add(new ValueParameter<SymbolicExpressionTreeCreator>("SolutionCreator", "The operator which should be used to create new symbolic regression solutions.", creator));
160      Parameters.Add(new ValueParameter<ISymbolicExpressionTreeInterpreter>("SymbolicExpressionTreeInterpreter", "The interpreter that should be used to evaluate the symbolic expression tree.", interpreter));
161      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)));
162      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)));
163      Parameters.Add(new ValueParameter<ISymbolicExpressionGrammar>("FunctionTreeGrammar", "The grammar that should be used for symbolic regression models.", globalGrammar));
164      Parameters.Add(new ValueParameter<IntValue>("MaxExpressionLength", "Maximal length of the symbolic expression.", new IntValue(100)));
165      Parameters.Add(new ValueParameter<IntValue>("MaxExpressionDepth", "Maximal depth of the symbolic expression.", new IntValue(10)));
166      Parameters.Add(new ValueParameter<IntValue>("MaxFunctionDefiningBranches", "Maximal number of automatically defined functions.", (IntValue)new IntValue(0).AsReadOnly()));
167      Parameters.Add(new ValueParameter<IntValue>("MaxFunctionArguments", "Maximal number of arguments of automatically defined functions.", (IntValue)new IntValue(0).AsReadOnly()));
168
169      creator.SymbolicExpressionTreeParameter.ActualName = "SymbolicRegressionModel";
170
171      ParameterizeSolutionCreator();
172
173      UpdateGrammar();
174      UpdateEstimationLimits();
175      InitializeOperators();
176      RegisterParameterEvents();
177      RegisterParameterValueEvents();
178    }
179
180    public override IDeepCloneable Clone(Cloner cloner) {
181      SymbolicRegressionProblemBase clone = (SymbolicRegressionProblemBase)base.Clone(cloner);
182      clone.operators = operators.Select(x => (IOperator)cloner.Clone(x)).ToList();
183      clone.RegisterParameterEvents();
184      clone.RegisterParameterValueEvents();
185      return clone;
186    }
187
188    private void RegisterParameterValueEvents() {
189      MaxFunctionArgumentsParameter.ValueChanged += new EventHandler(ArchitectureParameter_ValueChanged);
190      MaxFunctionDefiningBranchesParameter.ValueChanged += new EventHandler(ArchitectureParameter_ValueChanged);
191      SolutionCreatorParameter.ValueChanged += new EventHandler(SolutionCreatorParameter_ValueChanged);
192    }
193
194    private void RegisterParameterEvents() {
195      MaxFunctionArgumentsParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
196      MaxFunctionDefiningBranchesParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
197      SolutionCreator.SymbolicExpressionTreeParameter.ActualNameChanged += new EventHandler(SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged);
198    }
199
200    #region event handling
201    protected override void OnDataAnalysisProblemChanged(EventArgs e) {
202      base.OnDataAnalysisProblemChanged(e);
203      // paritions could be changed
204      ParameterizeAnalyzers();
205      // input variables could have been changed
206      UpdateGrammar();
207      // estimation limits have to be recalculated
208      UpdateEstimationLimits();
209    }
210    protected virtual void OnArchitectureParameterChanged(EventArgs e) {
211      UpdateGrammar();
212    }
213    protected virtual void OnGrammarChanged(EventArgs e) { }
214    protected virtual void OnOperatorsChanged(EventArgs e) { RaiseOperatorsChanged(e); }
215    protected virtual void OnSolutionCreatorChanged(EventArgs e) {
216      SolutionCreator.SymbolicExpressionTreeParameter.ActualNameChanged += new EventHandler(SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged);
217      ParameterizeSolutionCreator();
218      OnSolutionParameterNameChanged(e);
219      RaiseSolutionCreatorChanged(e);
220    }
221
222    protected virtual void OnSolutionParameterNameChanged(EventArgs e) {
223      ParameterizeAnalyzers();
224      ParameterizeOperators();
225    }
226
227    protected virtual void OnEvaluatorChanged(EventArgs e) {
228      RaiseEvaluatorChanged(e);
229    }
230    #endregion
231
232    #region event handlers
233    private void SolutionCreatorParameter_ValueChanged(object sender, EventArgs e) {
234      OnSolutionCreatorChanged(e);
235    }
236    private void SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged(object sender, EventArgs e) {
237      OnSolutionParameterNameChanged(e);
238    }
239    private void ArchitectureParameter_ValueChanged(object sender, EventArgs e) {
240      MaxFunctionArgumentsParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
241      MaxFunctionDefiningBranchesParameter.Value.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
242      OnArchitectureParameterChanged(e);
243    }
244    private void ArchitectureParameterValue_ValueChanged(object sender, EventArgs e) {
245      OnArchitectureParameterChanged(e);
246    }
247    #endregion
248
249    #region Helpers
250    [StorableHook(HookType.AfterDeserialization)]
251    private void AfterDeserializationHook() {
252      // BackwardsCompatibility3.3
253      #region Backwards compatible code (remove with 3.4)
254      if (operators == null) InitializeOperators();
255      #endregion
256      RegisterParameterEvents();
257      RegisterParameterValueEvents();
258    }
259
260    protected void AddOperator(IOperator op) {
261      operators.Add(op);
262    }
263
264    private void UpdateGrammar() {
265      foreach (var varSymbol in FunctionTreeGrammar.Symbols.OfType<HeuristicLab.Problems.DataAnalysis.Symbolic.Symbols.Variable>()) {
266        varSymbol.VariableNames = DataAnalysisProblemData.InputVariables.CheckedItems.Select(x => x.Value.Value);
267      }
268      var globalGrammar = FunctionTreeGrammar as GlobalSymbolicExpressionGrammar;
269      if (globalGrammar != null) {
270        globalGrammar.MaxFunctionArguments = MaxFunctionArguments.Value;
271        globalGrammar.MaxFunctionDefinitions = MaxFunctionDefiningBranches.Value;
272      }
273    }
274
275    private void UpdateEstimationLimits() {
276      if (TrainingSamplesStart.Value < TrainingSamplesEnd.Value &&
277        DataAnalysisProblemData.Dataset.VariableNames.Contains(DataAnalysisProblemData.TargetVariable.Value)) {
278        var targetValues = DataAnalysisProblemData.Dataset.GetVariableValues(DataAnalysisProblemData.TargetVariable.Value, TrainingSamplesStart.Value, TrainingSamplesEnd.Value);
279        var mean = targetValues.Average();
280        var range = targetValues.Max() - targetValues.Min();
281        UpperEstimationLimit = new DoubleValue(mean + PunishmentFactor.Value * range);
282        LowerEstimationLimit = new DoubleValue(mean - PunishmentFactor.Value * range);
283      }
284    }
285
286    private void InitializeOperators() {
287      operators = new List<IOperator>();
288      operators.AddRange(ApplicationManager.Manager.GetInstances<ISymbolicExpressionTreeOperator>().OfType<IOperator>());
289      operators.Add(new SymbolicRegressionTournamentPruning());
290      operators.Add(new SymbolicRegressionVariableFrequencyAnalyzer());
291      operators.Add(new MinAverageMaxSymbolicExpressionTreeSizeAnalyzer());
292      operators.Add(new SymbolicRegressionModelQualityAnalyzer());
293      ParameterizeOperators();
294      ParameterizeAnalyzers();
295    }
296
297    private void ParameterizeSolutionCreator() {
298      SolutionCreator.SymbolicExpressionGrammarParameter.ActualName = FunctionTreeGrammarParameter.Name;
299      SolutionCreator.MaxTreeHeightParameter.ActualName = MaxExpressionDepthParameter.Name;
300      SolutionCreator.MaxTreeSizeParameter.ActualName = MaxExpressionLengthParameter.Name;
301      SolutionCreator.MaxFunctionArgumentsParameter.ActualName = MaxFunctionArgumentsParameter.Name;
302      SolutionCreator.MaxFunctionDefinitionsParameter.ActualName = MaxFunctionDefiningBranchesParameter.Name;
303    }
304
305    private void ParameterizeAnalyzers() {
306      foreach (var analyzer in Analyzers) {
307        analyzer.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
308        var symbolicRegressionModelQualityAnalyzer = analyzer as SymbolicRegressionModelQualityAnalyzer;
309        if (symbolicRegressionModelQualityAnalyzer != null) {
310          symbolicRegressionModelQualityAnalyzer.ProblemDataParameter.ActualName = DataAnalysisProblemDataParameter.Name;
311          symbolicRegressionModelQualityAnalyzer.UpperEstimationLimitParameter.ActualName = UpperEstimationLimitParameter.Name;
312          symbolicRegressionModelQualityAnalyzer.LowerEstimationLimitParameter.ActualName = LowerEstimationLimitParameter.Name;
313          symbolicRegressionModelQualityAnalyzer.SymbolicExpressionTreeInterpreterParameter.ActualName = SymbolicExpressionTreeInterpreterParameter.Name;
314          symbolicRegressionModelQualityAnalyzer.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
315        }
316        var varFreqAnalyzer = analyzer as SymbolicRegressionVariableFrequencyAnalyzer;
317        if (varFreqAnalyzer != null) {
318          varFreqAnalyzer.ProblemDataParameter.ActualName = DataAnalysisProblemDataParameter.Name;
319        }
320        var pruningOperator = analyzer as SymbolicRegressionTournamentPruning;
321        if (pruningOperator != null) {
322          pruningOperator.SamplesStartParameter.Value = TrainingSamplesStart;
323          pruningOperator.SamplesEndParameter.Value = TrainingSamplesEnd;
324          pruningOperator.DataAnalysisProblemDataParameter.ActualName = DataAnalysisProblemDataParameter.Name;
325          pruningOperator.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
326          pruningOperator.SymbolicExpressionTreeInterpreterParameter.ActualName = SymbolicExpressionTreeInterpreterParameter.Name;
327          pruningOperator.LowerEstimationLimitParameter.ActualName = LowerEstimationLimitParameter.Name;
328          pruningOperator.UpperEstimationLimitParameter.ActualName = UpperEstimationLimitParameter.Name;
329        }
330      }
331      foreach (ISymbolicExpressionTreeAnalyzer analyzer in Operators.OfType<ISymbolicExpressionTreeAnalyzer>()) {
332        analyzer.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
333      }
334    }
335
336    private void ParameterizeOperators() {
337      foreach (ISymbolicExpressionTreeOperator op in Operators.OfType<ISymbolicExpressionTreeOperator>()) {
338        op.MaxTreeHeightParameter.ActualName = MaxExpressionDepthParameter.Name;
339        op.MaxTreeSizeParameter.ActualName = MaxExpressionLengthParameter.Name;
340        op.SymbolicExpressionGrammarParameter.ActualName = FunctionTreeGrammarParameter.Name;
341      }
342      foreach (ISymbolicExpressionTreeCrossover op in Operators.OfType<ISymbolicExpressionTreeCrossover>()) {
343        op.ParentsParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
344        op.ChildParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
345      }
346      foreach (ISymbolicExpressionTreeManipulator op in Operators.OfType<ISymbolicExpressionTreeManipulator>()) {
347        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
348      }
349      foreach (ISymbolicExpressionTreeArchitectureManipulator op in Operators.OfType<ISymbolicExpressionTreeArchitectureManipulator>()) {
350        op.MaxFunctionArgumentsParameter.ActualName = MaxFunctionArgumentsParameter.Name;
351        op.MaxFunctionDefinitionsParameter.ActualName = MaxFunctionDefiningBranchesParameter.Name;
352      }
353    }
354    #endregion
355  }
356}
Note: See TracBrowser for help on using the repository browser.