Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/SymbolicDataAnalysisProblem.cs @ 17753

Last change on this file since 17753 was 17413, checked in by mkommend, 5 years ago

#3052: Added DatanAnalyisGrammar for automatic configuration of variable symbols.
File size: 20.4 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 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 HEAL.Attic;
27using HeuristicLab.Common;
28using HeuristicLab.Common.Resources;
29using HeuristicLab.Core;
30using HeuristicLab.Data;
31using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
32using HeuristicLab.Optimization;
33using HeuristicLab.Parameters;
34using HeuristicLab.PluginInfrastructure;
35using HeuristicLab.Problems.Instances;
36
37namespace HeuristicLab.Problems.DataAnalysis.Symbolic {
38  [StorableType("59935E69-C4A5-480E-8FFB-D9669DE9BFD4")]
39  public abstract class SymbolicDataAnalysisProblem<T, U, V> : HeuristicOptimizationProblem<U, V>, IDataAnalysisProblem<T>, ISymbolicDataAnalysisProblem, IStorableContent,
40    IProblemInstanceConsumer<T>, IProblemInstanceExporter<T>
41    where T : class, IDataAnalysisProblemData
42    where U : class, ISymbolicDataAnalysisEvaluator<T>
43    where V : class, ISymbolicDataAnalysisSolutionCreator {
44
45    #region parameter names & descriptions
46    private const string ProblemDataParameterName = "ProblemData";
47    private const string SymbolicExpressionTreeGrammarParameterName = "SymbolicExpressionTreeGrammar";
48    private const string SymbolicExpressionTreeInterpreterParameterName = "SymbolicExpressionTreeInterpreter";
49    private const string MaximumSymbolicExpressionTreeDepthParameterName = "MaximumSymbolicExpressionTreeDepth";
50    private const string MaximumSymbolicExpressionTreeLengthParameterName = "MaximumSymbolicExpressionTreeLength";
51    private const string MaximumFunctionDefinitionsParameterName = "MaximumFunctionDefinitions";
52    private const string MaximumFunctionArgumentsParameterName = "MaximumFunctionArguments";
53    private const string RelativeNumberOfEvaluatedSamplesParameterName = "RelativeNumberOfEvaluatedSamples";
54    private const string FitnessCalculationPartitionParameterName = "FitnessCalculationPartition";
55    private const string ValidationPartitionParameterName = "ValidationPartition";
56    private const string ApplyLinearScalingParameterName = "ApplyLinearScaling";
57
58    private const string ProblemDataParameterDescription = "";
59    private const string SymbolicExpressionTreeGrammarParameterDescription = "The grammar that should be used for symbolic expression tree.";
60    private const string SymoblicExpressionTreeInterpreterParameterDescription = "The interpreter that should be used to evaluate the symbolic expression tree.";
61    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.";
62    private const string MaximumSymbolicExpressionTreeLengthParameterDescription = "Maximal length of the symbolic expression.";
63    private const string MaximumFunctionDefinitionsParameterDescription = "Maximal number of automatically defined functions";
64    private const string MaximumFunctionArgumentsParameterDescription = "Maximal number of arguments of automatically defined functions.";
65    private const string RelativeNumberOfEvaluatedSamplesParameterDescription = "The relative number of samples of the dataset partition, which should be randomly chosen for evaluation.";
66    private const string FitnessCalculationPartitionParameterDescription = "The partition of the problem data training partition, that should be used to calculate the fitness of an individual.";
67    private const string ValidationPartitionParameterDescription = "The partition of the problem data training partition, that should be used to select the best model from (optional).";
68    private const string ApplyLinearScalingParameterDescription = "Flag that indicates if the individual should be linearly scaled before evaluating.";
69    #endregion
70
71    #region parameter properties
72    IParameter IDataAnalysisProblem.ProblemDataParameter {
73      get { return ProblemDataParameter; }
74    }
75    public IValueParameter<T> ProblemDataParameter {
76      get { return (IValueParameter<T>)Parameters[ProblemDataParameterName]; }
77    }
78    public IValueParameter<ISymbolicDataAnalysisGrammar> SymbolicExpressionTreeGrammarParameter {
79      get { return (IValueParameter<ISymbolicDataAnalysisGrammar>)Parameters[SymbolicExpressionTreeGrammarParameterName]; }
80    }
81    public IValueParameter<ISymbolicDataAnalysisExpressionTreeInterpreter> SymbolicExpressionTreeInterpreterParameter {
82      get { return (IValueParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>)Parameters[SymbolicExpressionTreeInterpreterParameterName]; }
83    }
84    public IFixedValueParameter<IntValue> MaximumSymbolicExpressionTreeDepthParameter {
85      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumSymbolicExpressionTreeDepthParameterName]; }
86    }
87    public IFixedValueParameter<IntValue> MaximumSymbolicExpressionTreeLengthParameter {
88      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumSymbolicExpressionTreeLengthParameterName]; }
89    }
90    public IFixedValueParameter<IntValue> MaximumFunctionDefinitionsParameter {
91      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumFunctionDefinitionsParameterName]; }
92    }
93    public IFixedValueParameter<IntValue> MaximumFunctionArgumentsParameter {
94      get { return (IFixedValueParameter<IntValue>)Parameters[MaximumFunctionArgumentsParameterName]; }
95    }
96    public IFixedValueParameter<PercentValue> RelativeNumberOfEvaluatedSamplesParameter {
97      get { return (IFixedValueParameter<PercentValue>)Parameters[RelativeNumberOfEvaluatedSamplesParameterName]; }
98    }
99    public IFixedValueParameter<IntRange> FitnessCalculationPartitionParameter {
100      get { return (IFixedValueParameter<IntRange>)Parameters[FitnessCalculationPartitionParameterName]; }
101    }
102    public IFixedValueParameter<IntRange> ValidationPartitionParameter {
103      get { return (IFixedValueParameter<IntRange>)Parameters[ValidationPartitionParameterName]; }
104    }
105    public IFixedValueParameter<BoolValue> ApplyLinearScalingParameter {
106      get { return (IFixedValueParameter<BoolValue>)Parameters[ApplyLinearScalingParameterName]; }
107    }
108    #endregion
109
110    #region properties
111    public string Filename { get; set; }
112    public static new Image StaticItemImage { get { return VSImageLibrary.Type; } }
113
114    IDataAnalysisProblemData IDataAnalysisProblem.ProblemData {
115      get { return ProblemData; }
116    }
117    public T ProblemData {
118      get { return ProblemDataParameter.Value; }
119      set { ProblemDataParameter.Value = value; }
120    }
121
122    public ISymbolicDataAnalysisGrammar SymbolicExpressionTreeGrammar {
123      get { return SymbolicExpressionTreeGrammarParameter.Value; }
124      set { SymbolicExpressionTreeGrammarParameter.Value = value; }
125    }
126    public ISymbolicDataAnalysisExpressionTreeInterpreter SymbolicExpressionTreeInterpreter {
127      get { return SymbolicExpressionTreeInterpreterParameter.Value; }
128      set { SymbolicExpressionTreeInterpreterParameter.Value = value; }
129    }
130
131    public IntValue MaximumSymbolicExpressionTreeDepth {
132      get { return MaximumSymbolicExpressionTreeDepthParameter.Value; }
133    }
134    public IntValue MaximumSymbolicExpressionTreeLength {
135      get { return MaximumSymbolicExpressionTreeLengthParameter.Value; }
136    }
137    public IntValue MaximumFunctionDefinitions {
138      get { return MaximumFunctionDefinitionsParameter.Value; }
139    }
140    public IntValue MaximumFunctionArguments {
141      get { return MaximumFunctionArgumentsParameter.Value; }
142    }
143    public PercentValue RelativeNumberOfEvaluatedSamples {
144      get { return RelativeNumberOfEvaluatedSamplesParameter.Value; }
145    }
146
147    public IntRange FitnessCalculationPartition {
148      get { return FitnessCalculationPartitionParameter.Value; }
149    }
150    public IntRange ValidationPartition {
151      get { return ValidationPartitionParameter.Value; }
152    }
153    public BoolValue ApplyLinearScaling {
154      get { return ApplyLinearScalingParameter.Value; }
155    }
156    #endregion
157
158    [StorableConstructor]
159    protected SymbolicDataAnalysisProblem(StorableConstructorFlag _) : base(_) { }
160    [StorableHook(HookType.AfterDeserialization)]
161    private void AfterDeserialization() {
162      if (!Parameters.ContainsKey(ApplyLinearScalingParameterName)) {
163        Parameters.Add(new FixedValueParameter<BoolValue>(ApplyLinearScalingParameterName, ApplyLinearScalingParameterDescription, new BoolValue(false)));
164        ApplyLinearScalingParameter.Hidden = true;
165
166        //it is assumed that for all symbolic regression algorithms linear scaling was set to true
167        //there is no possibility to determine the previous value of the parameter as it was stored in the evaluator
168        if (GetType().Name.Contains("SymbolicRegression"))
169          ApplyLinearScaling.Value = true;
170      }
171
172      RegisterEventHandlers();
173    }
174    protected SymbolicDataAnalysisProblem(SymbolicDataAnalysisProblem<T, U, V> original, Cloner cloner)
175      : base(original, cloner) {
176      RegisterEventHandlers();
177    }
178
179    protected SymbolicDataAnalysisProblem(T problemData, U evaluator, V solutionCreator)
180      : base(evaluator, solutionCreator) {
181      Parameters.Add(new ValueParameter<T>(ProblemDataParameterName, ProblemDataParameterDescription, problemData));
182      Parameters.Add(new ValueParameter<ISymbolicDataAnalysisGrammar>(SymbolicExpressionTreeGrammarParameterName, SymbolicExpressionTreeGrammarParameterDescription));
183      Parameters.Add(new ValueParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>(SymbolicExpressionTreeInterpreterParameterName, SymoblicExpressionTreeInterpreterParameterDescription));
184      Parameters.Add(new FixedValueParameter<IntValue>(MaximumSymbolicExpressionTreeDepthParameterName, MaximumSymbolicExpressionTreeDepthParameterDescription));
185      Parameters.Add(new FixedValueParameter<IntValue>(MaximumSymbolicExpressionTreeLengthParameterName, MaximumSymbolicExpressionTreeLengthParameterDescription));
186      Parameters.Add(new FixedValueParameter<IntValue>(MaximumFunctionDefinitionsParameterName, MaximumFunctionDefinitionsParameterDescription));
187      Parameters.Add(new FixedValueParameter<IntValue>(MaximumFunctionArgumentsParameterName, MaximumFunctionArgumentsParameterDescription));
188      Parameters.Add(new FixedValueParameter<IntRange>(FitnessCalculationPartitionParameterName, FitnessCalculationPartitionParameterDescription));
189      Parameters.Add(new FixedValueParameter<IntRange>(ValidationPartitionParameterName, ValidationPartitionParameterDescription));
190      Parameters.Add(new FixedValueParameter<PercentValue>(RelativeNumberOfEvaluatedSamplesParameterName, RelativeNumberOfEvaluatedSamplesParameterDescription, new PercentValue(1)));
191      Parameters.Add(new FixedValueParameter<BoolValue>(ApplyLinearScalingParameterName, ApplyLinearScalingParameterDescription, new BoolValue(false)));
192
193      SymbolicExpressionTreeInterpreterParameter.Hidden = true;
194      MaximumFunctionArgumentsParameter.Hidden = true;
195      MaximumFunctionDefinitionsParameter.Hidden = true;
196      ApplyLinearScalingParameter.Hidden = true;
197
198      SymbolicExpressionTreeGrammar = new TypeCoherentExpressionGrammar();
199      SymbolicExpressionTreeInterpreter = new SymbolicDataAnalysisExpressionTreeLinearInterpreter();
200
201      FitnessCalculationPartition.Start = ProblemData.TrainingPartition.Start;
202      FitnessCalculationPartition.End = ProblemData.TrainingPartition.End;
203
204      InitializeOperators();
205
206      UpdateGrammar();
207      RegisterEventHandlers();
208    }
209
210    protected virtual void UpdateGrammar() {
211      var problemData = ProblemData;
212      var grammar = SymbolicExpressionTreeGrammar;
213
214      grammar.MaximumFunctionArguments = MaximumFunctionArguments.Value;
215      grammar.MaximumFunctionDefinitions = MaximumFunctionDefinitions.Value;
216
217      grammar.ConfigureVariableSymbols(problemData);
218    }
219
220    private void InitializeOperators() {
221      var operators = new HashSet<IItem>(new TypeEqualityComparer<IItem>());
222      operators.Add(new SubtreeCrossover());
223      operators.Add(new MultiSymbolicExpressionTreeManipulator());
224
225      foreach (var op in ApplicationManager.Manager.GetInstances<ISymbolicExpressionTreeOperator>())
226        operators.Add(op);
227      foreach (var op in ApplicationManager.Manager.GetInstances<ISymbolicDataAnalysisExpressionCrossover<T>>())
228        operators.Add(op);
229
230      operators.Add(new SymbolicExpressionSymbolFrequencyAnalyzer());
231      operators.Add(new SymbolicDataAnalysisVariableFrequencyAnalyzer());
232      operators.Add(new MinAverageMaxSymbolicExpressionTreeLengthAnalyzer());
233      operators.Add(new SymbolicExpressionTreeLengthAnalyzer());
234      operators.Add(new SymbolicExpressionTreeBottomUpSimilarityCalculator());
235      operators.Add(new SymbolicDataAnalysisBottomUpDiversityAnalyzer(operators.OfType<SymbolicExpressionTreeBottomUpSimilarityCalculator>().First()));
236
237      Operators.AddRange(operators);
238      ParameterizeOperators();
239    }
240
241    #region events
242    private void RegisterEventHandlers() {
243      ProblemDataParameter.ValueChanged += new EventHandler(ProblemDataParameter_ValueChanged);
244      ProblemDataParameter.Value.Changed += (object sender, EventArgs e) => OnProblemDataChanged();
245
246      SymbolicExpressionTreeGrammarParameter.ValueChanged += new EventHandler(SymbolicExpressionTreeGrammarParameter_ValueChanged);
247
248      MaximumFunctionArguments.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
249      MaximumFunctionDefinitions.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
250      MaximumSymbolicExpressionTreeDepth.ValueChanged += new EventHandler(MaximumSymbolicExpressionTreeDepth_ValueChanged);
251    }
252
253    private void ProblemDataParameter_ValueChanged(object sender, EventArgs e) {
254      ValidationPartition.Start = 0;
255      ValidationPartition.End = 0;
256      ProblemDataParameter.Value.Changed += (object s, EventArgs args) => OnProblemDataChanged();
257      OnProblemDataChanged();
258    }
259
260    private void SymbolicExpressionTreeGrammarParameter_ValueChanged(object sender, EventArgs e) {
261      UpdateGrammar();
262    }
263
264    private void ArchitectureParameterValue_ValueChanged(object sender, EventArgs e) {
265      UpdateGrammar();
266    }
267
268    private void MaximumSymbolicExpressionTreeDepth_ValueChanged(object sender, EventArgs e) {
269      if (MaximumSymbolicExpressionTreeDepth != null && MaximumSymbolicExpressionTreeDepth.Value < 3)
270        MaximumSymbolicExpressionTreeDepth.Value = 3;
271    }
272
273    protected override void OnSolutionCreatorChanged() {
274      base.OnSolutionCreatorChanged();
275      SolutionCreator.SymbolicExpressionTreeParameter.ActualNameChanged += new EventHandler(SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged);
276      ParameterizeOperators();
277    }
278
279    private void SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged(object sender, EventArgs e) {
280      ParameterizeOperators();
281    }
282
283    protected override void OnEvaluatorChanged() {
284      base.OnEvaluatorChanged();
285      ParameterizeOperators();
286    }
287
288    public event EventHandler ProblemDataChanged;
289    protected virtual void OnProblemDataChanged() {
290      FitnessCalculationPartition.Start = ProblemData.TrainingPartition.Start;
291      FitnessCalculationPartition.End = ProblemData.TrainingPartition.End;
292
293      UpdateGrammar();
294      ParameterizeOperators();
295
296      var handler = ProblemDataChanged;
297      if (handler != null) handler(this, EventArgs.Empty);
298
299      OnReset();
300    }
301    #endregion
302
303    protected virtual void ParameterizeOperators() {
304      var operators = Parameters.OfType<IValueParameter>().Select(p => p.Value).OfType<IOperator>().Union(Operators).ToList();
305
306      foreach (var op in operators.OfType<ISymbolicExpressionTreeGrammarBasedOperator>()) {
307        op.SymbolicExpressionTreeGrammarParameter.ActualName = SymbolicExpressionTreeGrammarParameter.Name;
308      }
309      foreach (var op in operators.OfType<ISymbolicExpressionTreeSizeConstraintOperator>()) {
310        op.MaximumSymbolicExpressionTreeDepthParameter.ActualName = MaximumSymbolicExpressionTreeDepthParameter.Name;
311        op.MaximumSymbolicExpressionTreeLengthParameter.ActualName = MaximumSymbolicExpressionTreeLengthParameter.Name;
312      }
313      foreach (var op in operators.OfType<ISymbolicExpressionTreeArchitectureAlteringOperator>()) {
314        op.MaximumFunctionArgumentsParameter.ActualName = MaximumFunctionArgumentsParameter.Name;
315        op.MaximumFunctionDefinitionsParameter.ActualName = MaximumFunctionDefinitionsParameter.Name;
316      }
317      foreach (var op in operators.OfType<ISymbolicDataAnalysisEvaluator<T>>()) {
318        op.ProblemDataParameter.ActualName = ProblemDataParameterName;
319        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
320        op.EvaluationPartitionParameter.ActualName = FitnessCalculationPartitionParameter.Name;
321        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
322        op.ApplyLinearScalingParameter.ActualName = ApplyLinearScalingParameter.Name;
323      }
324      foreach (var op in operators.OfType<ISymbolicExpressionTreeCrossover>()) {
325        op.ParentsParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
326        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
327      }
328      foreach (var op in operators.OfType<ISymbolicExpressionTreeManipulator>()) {
329        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
330      }
331      foreach (var op in operators.OfType<ISymbolicExpressionTreeAnalyzer>()) {
332        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
333      }
334      foreach (var op in operators.OfType<ISymbolicDataAnalysisSingleObjectiveAnalyzer>()) {
335        op.ApplyLinearScalingParameter.ActualName = ApplyLinearScalingParameter.Name;
336      }
337      foreach (var op in operators.OfType<ISymbolicDataAnalysisMultiObjectiveAnalyzer>()) {
338        op.ApplyLinearScalingParameter.ActualName = ApplyLinearScalingParameter.Name;
339      }
340      foreach (var op in operators.OfType<ISymbolicDataAnalysisAnalyzer>()) {
341        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
342      }
343      foreach (var op in operators.OfType<ISymbolicDataAnalysisValidationAnalyzer<U, T>>()) {
344        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
345        op.ValidationPartitionParameter.ActualName = ValidationPartitionParameter.Name;
346      }
347      foreach (var op in operators.OfType<ISymbolicDataAnalysisInterpreterOperator>()) {
348        op.SymbolicDataAnalysisTreeInterpreterParameter.ActualName = SymbolicExpressionTreeInterpreterParameter.Name;
349      }
350      foreach (var op in operators.OfType<ISymbolicDataAnalysisExpressionCrossover<T>>()) {
351        op.EvaluationPartitionParameter.ActualName = FitnessCalculationPartitionParameter.Name;
352        op.ProblemDataParameter.ActualName = ProblemDataParameter.Name;
353        op.EvaluationPartitionParameter.ActualName = FitnessCalculationPartitionParameter.Name;
354        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
355        op.EvaluatorParameter.ActualName = EvaluatorParameter.Name;
356      }
357    }
358
359    #region Import & Export
360    public virtual void Load(T data) {
361      Name = data.Name;
362      Description = data.Description;
363      ProblemData = data;
364    }
365
366    public virtual T Export() {
367      return ProblemData;
368    }
369    #endregion
370  }
371}
Note: See TracBrowser for help on using the repository browser.