Free cookie consent management tool by TermsFeed Policy Generator

source: branches/RegressionBenchmarks/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/SymbolicDataAnalysisProblem.cs @ 7025

Last change on this file since 7025 was 7025, checked in by sforsten, 13 years ago

#1669: benchmark problems of Nguyen, Korns and Keijzer from ​http://groups.csail.mit.edu/EVO-DesignOpt/GPBenchmarks/ have been added. The benchmark problems from ​http://www.vanillamodeling.com/ have been adapted to the ones from Vladislavleva.

Not all benchmarks are working correctly so far, but they will be tested soon.
File size: 17.8 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2011 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.Drawing;
25using System.Linq;
26using HeuristicLab.Common;
27using HeuristicLab.Common.Resources;
28using HeuristicLab.Core;
29using HeuristicLab.Data;
30using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
31using HeuristicLab.Optimization;
32using HeuristicLab.Parameters;
33using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
34using HeuristicLab.PluginInfrastructure;
35
36namespace HeuristicLab.Problems.DataAnalysis.Symbolic {
37  [StorableClass]
38  public abstract class SymbolicDataAnalysisProblem<T, U, V> : HeuristicOptimizationProblem<U, V>, IDataAnalysisProblem<T>, ISymbolicDataAnalysisProblem, IStorableContent
39    where T : class,IDataAnalysisProblemData
40    where U : class, ISymbolicDataAnalysisEvaluator<T>
41    where V : class, ISymbolicDataAnalysisSolutionCreator {
42
43    #region parameter names & descriptions
44    private const string ProblemDataParameterName = "ProblemData";
45    private const string SymbolicExpressionTreeGrammarParameterName = "SymbolicExpressionTreeGrammar";
46    private const string SymbolicExpressionTreeInterpreterParameterName = "SymbolicExpressionTreeInterpreter";
47    private const string MaximumSymbolicExpressionTreeDepthParameterName = "MaximumSymbolicExpressionTreeDepth";
48    private const string MaximumSymbolicExpressionTreeLengthParameterName = "MaximumSymbolicExpressionTreeLength";
49    private const string MaximumFunctionDefinitionsParameterName = "MaximumFunctionDefinitions";
50    private const string MaximumFunctionArgumentsParameterName = "MaximumFunctionArguments";
51    private const string RelativeNumberOfEvaluatedSamplesParameterName = "RelativeNumberOfEvaluatedSamples";
52    private const string FitnessCalculationPartitionParameterName = "FitnessCalculationPartition";
53    private const string ValidationPartitionParameterName = "ValidationPartition";
54
55    private const string ProblemDataParameterDescription = "";
56    private const string SymbolicExpressionTreeGrammarParameterDescription = "The grammar that should be used for symbolic expression tree.";
57    private const string SymoblicExpressionTreeInterpreterParameterDescription = "The interpreter that should be used to evaluate the symbolic expression tree.";
58    private const string MaximumSymbolicExpressionTreeDepthParameterDescription = "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.";
59    private const string MaximumSymbolicExpressionTreeLengthParameterDescription = "Maximal length of the symbolic expression.";
60    private const string MaximumFunctionDefinitionsParameterDescription = "Maximal number of automatically defined functions";
61    private const string MaximumFunctionArgumentsParameterDescription = "Maximal number of arguments of automatically defined functions.";
62    private const string RelativeNumberOfEvaluatedSamplesParameterDescription = "The relative number of samples of the dataset partition, which should be randomly chosen for evaluation.";
63    private const string FitnessCalculationPartitionParameterDescription = "The partition of the problem data training partition, that should be used to calculate the fitness of an individual.";
64    private const string ValidationPartitionParameterDescription = "The partition of the problem data training partition, that should be used to select the best model from (optional).";
65    #endregion
66
67    #region parameter properties
68    IParameter IDataAnalysisProblem.ProblemDataParameter {
69      get { return ProblemDataParameter; }
70    }
71    public IValueParameter<T> ProblemDataParameter {
72      get { return (IValueParameter<T>)Parameters[ProblemDataParameterName]; }
73    }
74    public IValueParameter<ISymbolicDataAnalysisGrammar> SymbolicExpressionTreeGrammarParameter {
75      get { return (IValueParameter<ISymbolicDataAnalysisGrammar>)Parameters[SymbolicExpressionTreeGrammarParameterName]; }
76    }
77    public IValueParameter<ISymbolicDataAnalysisExpressionTreeInterpreter> SymbolicExpressionTreeInterpreterParameter {
78      get { return (IValueParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>)Parameters[SymbolicExpressionTreeInterpreterParameterName]; }
79    }
80    public IFixedValueParameter<IntValue> MaximumSymbolicExpressionTreeDepthParameter {
81      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumSymbolicExpressionTreeDepthParameterName]; }
82    }
83    public IFixedValueParameter<IntValue> MaximumSymbolicExpressionTreeLengthParameter {
84      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumSymbolicExpressionTreeLengthParameterName]; }
85    }
86    public IFixedValueParameter<IntValue> MaximumFunctionDefinitionsParameter {
87      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumFunctionDefinitionsParameterName]; }
88    }
89    public IFixedValueParameter<IntValue> MaximumFunctionArgumentsParameter {
90      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumFunctionArgumentsParameterName]; }
91    }
92    public IFixedValueParameter<PercentValue> RelativeNumberOfEvaluatedSamplesParameter {
93      get { return (IFixedValueParameter<PercentValue>)Parameters[RelativeNumberOfEvaluatedSamplesParameterName]; }
94    }
95    public IFixedValueParameter<IntRange> FitnessCalculationPartitionParameter {
96      get { return (IFixedValueParameter<IntRange>)Parameters[FitnessCalculationPartitionParameterName]; }
97    }
98    public IFixedValueParameter<IntRange> ValidationPartitionParameter {
99      get { return (IFixedValueParameter<IntRange>)Parameters[ValidationPartitionParameterName]; }
100    }
101    #endregion
102
103    #region properties
104    public string Filename { get; set; }
105    public override Image ItemImage { get { return VSImageLibrary.Type; } }
106
107    IDataAnalysisProblemData IDataAnalysisProblem.ProblemData {
108      get { return ProblemData; }
109      set { ProblemData = (T)value; }
110
111    }
112    public T ProblemData {
113      get { return ProblemDataParameter.Value; }
114      set { ProblemDataParameter.Value = value; }
115    }
116
117    public ISymbolicDataAnalysisGrammar SymbolicExpressionTreeGrammar {
118      get { return SymbolicExpressionTreeGrammarParameter.Value; }
119      set { SymbolicExpressionTreeGrammarParameter.Value = value; }
120    }
121    public ISymbolicDataAnalysisExpressionTreeInterpreter SymbolicExpressionTreeInterpreter {
122      get { return SymbolicExpressionTreeInterpreterParameter.Value; }
123      set { SymbolicExpressionTreeInterpreterParameter.Value = value; }
124    }
125
126    public IntValue MaximumSymbolicExpressionTreeDepth {
127      get { return MaximumSymbolicExpressionTreeDepthParameter.Value; }
128    }
129    public IntValue MaximumSymbolicExpressionTreeLength {
130      get { return MaximumSymbolicExpressionTreeLengthParameter.Value; }
131    }
132    public IntValue MaximumFunctionDefinitions {
133      get { return MaximumFunctionDefinitionsParameter.Value; }
134    }
135    public IntValue MaximumFunctionArguments {
136      get { return MaximumFunctionArgumentsParameter.Value; }
137    }
138    public PercentValue RelativeNumberOfEvaluatedSamples {
139      get { return RelativeNumberOfEvaluatedSamplesParameter.Value; }
140    }
141
142    public IntRange FitnessCalculationPartition {
143      get { return FitnessCalculationPartitionParameter.Value; }
144    }
145    public IntRange ValidationPartition {
146      get { return ValidationPartitionParameter.Value; }
147    }
148    #endregion
149
150    [StorableConstructor]
151    protected SymbolicDataAnalysisProblem(bool deserializing) : base(deserializing) { }
152    [StorableHook(HookType.AfterDeserialization)]
153    private void AfterDeserialization() {
154      RegisterEventHandlers();
155    }
156    protected SymbolicDataAnalysisProblem(SymbolicDataAnalysisProblem<T, U, V> original, Cloner cloner)
157      : base(original, cloner) {
158      RegisterEventHandlers();
159    }
160
161    protected SymbolicDataAnalysisProblem(T problemData, U evaluator, V solutionCreator)
162      : base(evaluator, solutionCreator) {
163      Parameters.Add(new ValueParameter<T>(ProblemDataParameterName, ProblemDataParameterDescription, problemData));
164      Parameters.Add(new ValueParameter<ISymbolicDataAnalysisGrammar>(SymbolicExpressionTreeGrammarParameterName, SymbolicExpressionTreeGrammarParameterDescription));
165      Parameters.Add(new ValueParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>(SymbolicExpressionTreeInterpreterParameterName, SymoblicExpressionTreeInterpreterParameterDescription));
166      Parameters.Add(new FixedValueParameter<IntValue>(MaximumSymbolicExpressionTreeDepthParameterName, MaximumSymbolicExpressionTreeDepthParameterDescription));
167      Parameters.Add(new FixedValueParameter<IntValue>(MaximumSymbolicExpressionTreeLengthParameterName, MaximumSymbolicExpressionTreeLengthParameterDescription));
168      Parameters.Add(new FixedValueParameter<IntValue>(MaximumFunctionDefinitionsParameterName, MaximumFunctionDefinitionsParameterDescription));
169      Parameters.Add(new FixedValueParameter<IntValue>(MaximumFunctionArgumentsParameterName, MaximumFunctionArgumentsParameterDescription));
170      Parameters.Add(new FixedValueParameter<IntRange>(FitnessCalculationPartitionParameterName, FitnessCalculationPartitionParameterDescription));
171      Parameters.Add(new FixedValueParameter<IntRange>(ValidationPartitionParameterName, ValidationPartitionParameterDescription));
172      Parameters.Add(new FixedValueParameter<PercentValue>(RelativeNumberOfEvaluatedSamplesParameterName, RelativeNumberOfEvaluatedSamplesParameterDescription, new PercentValue(1)));
173
174      SolutionCreatorParameter.Hidden = true;
175      SymbolicExpressionTreeInterpreterParameter.Hidden = true;
176      MaximumFunctionArgumentsParameter.Hidden = true;
177      MaximumFunctionDefinitionsParameter.Hidden = true;
178
179      SymbolicExpressionTreeGrammar = new TypeCoherentExpressionGrammar();
180      SymbolicExpressionTreeInterpreter = new SymbolicDataAnalysisExpressionTreeInterpreter();
181
182      FitnessCalculationPartition.Start = ProblemData.TrainingPartition.Start;
183      FitnessCalculationPartition.End = ProblemData.TrainingPartition.End;
184
185      InitializeOperators();
186
187      UpdateGrammar();
188      RegisterEventHandlers();
189    }
190
191    protected virtual void UpdateGrammar() {
192      SymbolicExpressionTreeGrammar.MaximumFunctionArguments = MaximumFunctionArguments.Value;
193      SymbolicExpressionTreeGrammar.MaximumFunctionDefinitions = MaximumFunctionDefinitions.Value;
194      foreach (var varSymbol in SymbolicExpressionTreeGrammar.Symbols.OfType<HeuristicLab.Problems.DataAnalysis.Symbolic.Variable>()) {
195        if (!varSymbol.Fixed) varSymbol.VariableNames = ProblemData.AllowedInputVariables;
196      }
197      foreach (var varSymbol in SymbolicExpressionTreeGrammar.Symbols.OfType<HeuristicLab.Problems.DataAnalysis.Symbolic.VariableCondition>()) {
198        if (!varSymbol.Fixed) varSymbol.VariableNames = ProblemData.AllowedInputVariables;
199      }
200    }
201
202    private void InitializeOperators() {
203      Operators.AddRange(ApplicationManager.Manager.GetInstances<ISymbolicExpressionTreeOperator>());
204      Operators.Add(new SymbolicExpressionSymbolFrequencyAnalyzer());
205      Operators.Add(new SymbolicDataAnalysisVariableFrequencyAnalyzer());
206      Operators.Add(new MinAverageMaxSymbolicExpressionTreeLengthAnalyzer());
207      ParameterizeOperators();
208    }
209
210    #region events
211    private void RegisterEventHandlers() {
212      ProblemDataParameter.ValueChanged += new EventHandler(ProblemDataParameter_ValueChanged);
213      ProblemDataParameter.Value.Changed += (object sender, EventArgs e) => OnProblemDataChanged();
214
215      SymbolicExpressionTreeGrammarParameter.ValueChanged += new EventHandler(SymbolicExpressionTreeGrammarParameter_ValueChanged);
216
217      MaximumFunctionArguments.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
218      MaximumFunctionDefinitions.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
219      MaximumSymbolicExpressionTreeDepth.ValueChanged += new EventHandler(MaximumSymbolicExpressionTreeDepth_ValueChanged);
220    }
221
222    private void ProblemDataParameter_ValueChanged(object sender, EventArgs e) {
223      ValidationPartition.Start = 0;
224      ValidationPartition.End = 0;
225      ProblemDataParameter.Value.Changed += (object s, EventArgs args) => OnProblemDataChanged();
226      OnProblemDataChanged();
227    }
228
229    private void SymbolicExpressionTreeGrammarParameter_ValueChanged(object sender, EventArgs e) {
230      UpdateGrammar();
231    }
232
233    private void ArchitectureParameterValue_ValueChanged(object sender, EventArgs e) {
234      UpdateGrammar();
235    }
236
237    private void MaximumSymbolicExpressionTreeDepth_ValueChanged(object sender, EventArgs e) {
238      if (MaximumSymbolicExpressionTreeDepth != null && MaximumSymbolicExpressionTreeDepth.Value < 3)
239        MaximumSymbolicExpressionTreeDepth.Value = 3;
240    }
241
242    protected override void OnSolutionCreatorChanged() {
243      base.OnSolutionCreatorChanged();
244      SolutionCreator.SymbolicExpressionTreeParameter.ActualNameChanged += new EventHandler(SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged);
245      ParameterizeOperators();
246    }
247
248    private void SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged(object sender, EventArgs e) {
249      ParameterizeOperators();
250    }
251
252    protected override void OnEvaluatorChanged() {
253      base.OnEvaluatorChanged();
254      ParameterizeOperators();
255    }
256
257    public event EventHandler ProblemDataChanged;
258    protected virtual void OnProblemDataChanged() {
259      FitnessCalculationPartition.Start = ProblemData.TrainingPartition.Start;
260      FitnessCalculationPartition.End = ProblemData.TrainingPartition.End;
261
262      UpdateGrammar();
263      ParameterizeOperators();
264
265      var handler = ProblemDataChanged;
266      if (handler != null) handler(this, EventArgs.Empty);
267
268      OnReset();
269    }
270    #endregion
271
272    protected virtual void ParameterizeOperators() {
273      var operators = Parameters.OfType<IValueParameter>().Select(p => p.Value).OfType<IOperator>().Union(Operators);
274
275      foreach (var op in operators.OfType<ISymbolicExpressionTreeGrammarBasedOperator>()) {
276        op.SymbolicExpressionTreeGrammarParameter.ActualName = SymbolicExpressionTreeGrammarParameterName;
277      }
278      foreach (var op in operators.OfType<ISymbolicExpressionTreeSizeConstraintOperator>()) {
279        op.MaximumSymbolicExpressionTreeDepthParameter.ActualName = MaximumSymbolicExpressionTreeDepthParameterName;
280        op.MaximumSymbolicExpressionTreeLengthParameter.ActualName = MaximumSymbolicExpressionTreeLengthParameterName;
281      }
282      foreach (var op in operators.OfType<ISymbolicExpressionTreeArchitectureAlteringOperator>()) {
283        op.MaximumFunctionArgumentsParameter.ActualName = MaximumFunctionArgumentsParameterName;
284        op.MaximumFunctionDefinitionsParameter.ActualName = MaximumFunctionDefinitionsParameterName;
285      }
286      foreach (var op in operators.OfType<ISymbolicDataAnalysisEvaluator<T>>()) {
287        op.ProblemDataParameter.ActualName = ProblemDataParameterName;
288        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
289        op.EvaluationPartitionParameter.ActualName = FitnessCalculationPartitionParameter.Name;
290        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
291      }
292      foreach (var op in operators.OfType<ISymbolicExpressionTreeCrossover>()) {
293        op.ParentsParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
294        op.ChildParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
295      }
296      foreach (var op in operators.OfType<ISymbolicExpressionTreeManipulator>()) {
297        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
298      }
299      foreach (var op in operators.OfType<ISymbolicExpressionTreeAnalyzer>()) {
300        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
301      }
302      foreach (var op in operators.OfType<ISymbolicDataAnalysisAnalyzer>()) {
303        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
304      }
305      foreach (var op in operators.OfType<ISymbolicDataAnalysisValidationAnalyzer<U, T>>()) {
306        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
307        op.ValidationPartitionParameter.ActualName = ValidationPartitionParameter.Name;
308      }
309      foreach (var op in operators.OfType<ISymbolicDataAnalysisInterpreterOperator>()) {
310        op.SymbolicDataAnalysisTreeInterpreterParameter.ActualName = SymbolicExpressionTreeInterpreterParameterName;
311      }
312    }
313
314    public abstract void ImportProblemDataFromFile(string fileName);
315
316    public abstract void CreateProblemDataFromBenchmark(IDataAnalysisBenchmarkProblemDataGenerator benchmarkGenerator);
317
318    public abstract IEnumerable<IDataAnalysisBenchmarkProblemDataGenerator> GetBenchmarkProblemDataGenerators();
319  }
320}
Note: See TracBrowser for help on using the repository browser.