Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/SymbolicDataAnalysisProblem.cs @ 16858

Last change on this file since 16858 was 16565, checked in by gkronber, 6 years ago

#2520: merged changes from PersistenceOverhaul branch (r16451:16564) into trunk
File size: 22.0 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2019 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 HEAL.Attic;
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 ds = problemData.Dataset;
213      var grammar = SymbolicExpressionTreeGrammar;
214      grammar.MaximumFunctionArguments = MaximumFunctionArguments.Value;
215      grammar.MaximumFunctionDefinitions = MaximumFunctionDefinitions.Value;
216      foreach (var varSymbol in grammar.Symbols.OfType<HeuristicLab.Problems.DataAnalysis.Symbolic.VariableBase>()) {
217        if (!varSymbol.Fixed) {
218          varSymbol.AllVariableNames = problemData.InputVariables.Select(x => x.Value).Where(x => ds.VariableHasType<double>(x));
219          varSymbol.VariableNames = problemData.AllowedInputVariables.Where(x => ds.VariableHasType<double>(x));
220        }
221      }
222      foreach (var factorSymbol in grammar.Symbols.OfType<BinaryFactorVariable>()) {
223        if (!factorSymbol.Fixed) {
224          factorSymbol.AllVariableNames = problemData.InputVariables.Select(x => x.Value).Where(x => ds.VariableHasType<string>(x));
225          factorSymbol.VariableNames = problemData.AllowedInputVariables.Where(x => ds.VariableHasType<string>(x));
226          factorSymbol.VariableValues = factorSymbol.VariableNames
227            .ToDictionary(varName => varName, varName => ds.GetStringValues(varName).Distinct().ToList());
228        }
229      }
230      foreach (var factorSymbol in grammar.Symbols.OfType<FactorVariable>()) {
231        if (!factorSymbol.Fixed) {
232          factorSymbol.AllVariableNames = problemData.InputVariables.Select(x => x.Value).Where(x => ds.VariableHasType<string>(x));
233          factorSymbol.VariableNames = problemData.AllowedInputVariables.Where(x => ds.VariableHasType<string>(x));
234          factorSymbol.VariableValues = factorSymbol.VariableNames
235            .ToDictionary(varName => varName,
236            varName => ds.GetStringValues(varName).Distinct()
237            .Select((n, i) => Tuple.Create(n, i))
238            .ToDictionary(tup => tup.Item1, tup => tup.Item2));
239        }
240      }
241    }
242
243    private void InitializeOperators() {
244      var operators = new HashSet<IItem>(new TypeEqualityComparer<IItem>());
245      operators.Add(new SubtreeCrossover());
246      operators.Add(new MultiSymbolicExpressionTreeManipulator());
247
248      foreach (var op in ApplicationManager.Manager.GetInstances<ISymbolicExpressionTreeOperator>())
249        operators.Add(op);
250      foreach (var op in ApplicationManager.Manager.GetInstances<ISymbolicDataAnalysisExpressionCrossover<T>>())
251        operators.Add(op);
252
253      operators.Add(new SymbolicExpressionSymbolFrequencyAnalyzer());
254      operators.Add(new SymbolicDataAnalysisVariableFrequencyAnalyzer());
255      operators.Add(new MinAverageMaxSymbolicExpressionTreeLengthAnalyzer());
256      operators.Add(new SymbolicExpressionTreeLengthAnalyzer());
257      operators.Add(new SymbolicExpressionTreeBottomUpSimilarityCalculator());
258      operators.Add(new SymbolicDataAnalysisBottomUpDiversityAnalyzer(operators.OfType<SymbolicExpressionTreeBottomUpSimilarityCalculator>().First()));
259
260      Operators.AddRange(operators);
261      ParameterizeOperators();
262    }
263
264    #region events
265    private void RegisterEventHandlers() {
266      ProblemDataParameter.ValueChanged += new EventHandler(ProblemDataParameter_ValueChanged);
267      ProblemDataParameter.Value.Changed += (object sender, EventArgs e) => OnProblemDataChanged();
268
269      SymbolicExpressionTreeGrammarParameter.ValueChanged += new EventHandler(SymbolicExpressionTreeGrammarParameter_ValueChanged);
270
271      MaximumFunctionArguments.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
272      MaximumFunctionDefinitions.ValueChanged += new EventHandler(ArchitectureParameterValue_ValueChanged);
273      MaximumSymbolicExpressionTreeDepth.ValueChanged += new EventHandler(MaximumSymbolicExpressionTreeDepth_ValueChanged);
274    }
275
276    private void ProblemDataParameter_ValueChanged(object sender, EventArgs e) {
277      ValidationPartition.Start = 0;
278      ValidationPartition.End = 0;
279      ProblemDataParameter.Value.Changed += (object s, EventArgs args) => OnProblemDataChanged();
280      OnProblemDataChanged();
281    }
282
283    private void SymbolicExpressionTreeGrammarParameter_ValueChanged(object sender, EventArgs e) {
284      UpdateGrammar();
285    }
286
287    private void ArchitectureParameterValue_ValueChanged(object sender, EventArgs e) {
288      UpdateGrammar();
289    }
290
291    private void MaximumSymbolicExpressionTreeDepth_ValueChanged(object sender, EventArgs e) {
292      if (MaximumSymbolicExpressionTreeDepth != null && MaximumSymbolicExpressionTreeDepth.Value < 3)
293        MaximumSymbolicExpressionTreeDepth.Value = 3;
294    }
295
296    protected override void OnSolutionCreatorChanged() {
297      base.OnSolutionCreatorChanged();
298      SolutionCreator.SymbolicExpressionTreeParameter.ActualNameChanged += new EventHandler(SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged);
299      ParameterizeOperators();
300    }
301
302    private void SolutionCreator_SymbolicExpressionTreeParameter_ActualNameChanged(object sender, EventArgs e) {
303      ParameterizeOperators();
304    }
305
306    protected override void OnEvaluatorChanged() {
307      base.OnEvaluatorChanged();
308      ParameterizeOperators();
309    }
310
311    public event EventHandler ProblemDataChanged;
312    protected virtual void OnProblemDataChanged() {
313      FitnessCalculationPartition.Start = ProblemData.TrainingPartition.Start;
314      FitnessCalculationPartition.End = ProblemData.TrainingPartition.End;
315
316      UpdateGrammar();
317      ParameterizeOperators();
318
319      var handler = ProblemDataChanged;
320      if (handler != null) handler(this, EventArgs.Empty);
321
322      OnReset();
323    }
324    #endregion
325
326    protected virtual void ParameterizeOperators() {
327      var operators = Parameters.OfType<IValueParameter>().Select(p => p.Value).OfType<IOperator>().Union(Operators).ToList();
328
329      foreach (var op in operators.OfType<ISymbolicExpressionTreeGrammarBasedOperator>()) {
330        op.SymbolicExpressionTreeGrammarParameter.ActualName = SymbolicExpressionTreeGrammarParameter.Name;
331      }
332      foreach (var op in operators.OfType<ISymbolicExpressionTreeSizeConstraintOperator>()) {
333        op.MaximumSymbolicExpressionTreeDepthParameter.ActualName = MaximumSymbolicExpressionTreeDepthParameter.Name;
334        op.MaximumSymbolicExpressionTreeLengthParameter.ActualName = MaximumSymbolicExpressionTreeLengthParameter.Name;
335      }
336      foreach (var op in operators.OfType<ISymbolicExpressionTreeArchitectureAlteringOperator>()) {
337        op.MaximumFunctionArgumentsParameter.ActualName = MaximumFunctionArgumentsParameter.Name;
338        op.MaximumFunctionDefinitionsParameter.ActualName = MaximumFunctionDefinitionsParameter.Name;
339      }
340      foreach (var op in operators.OfType<ISymbolicDataAnalysisEvaluator<T>>()) {
341        op.ProblemDataParameter.ActualName = ProblemDataParameterName;
342        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
343        op.EvaluationPartitionParameter.ActualName = FitnessCalculationPartitionParameter.Name;
344        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
345        op.ApplyLinearScalingParameter.ActualName = ApplyLinearScalingParameter.Name;
346      }
347      foreach (var op in operators.OfType<ISymbolicExpressionTreeCrossover>()) {
348        op.ParentsParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
349        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
350      }
351      foreach (var op in operators.OfType<ISymbolicExpressionTreeManipulator>()) {
352        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
353      }
354      foreach (var op in operators.OfType<ISymbolicExpressionTreeAnalyzer>()) {
355        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
356      }
357      foreach (var op in operators.OfType<ISymbolicDataAnalysisSingleObjectiveAnalyzer>()) {
358        op.ApplyLinearScalingParameter.ActualName = ApplyLinearScalingParameter.Name;
359      }
360      foreach (var op in operators.OfType<ISymbolicDataAnalysisMultiObjectiveAnalyzer>()) {
361        op.ApplyLinearScalingParameter.ActualName = ApplyLinearScalingParameter.Name;
362      }
363      foreach (var op in operators.OfType<ISymbolicDataAnalysisAnalyzer>()) {
364        op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
365      }
366      foreach (var op in operators.OfType<ISymbolicDataAnalysisValidationAnalyzer<U, T>>()) {
367        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
368        op.ValidationPartitionParameter.ActualName = ValidationPartitionParameter.Name;
369      }
370      foreach (var op in operators.OfType<ISymbolicDataAnalysisInterpreterOperator>()) {
371        op.SymbolicDataAnalysisTreeInterpreterParameter.ActualName = SymbolicExpressionTreeInterpreterParameter.Name;
372      }
373      foreach (var op in operators.OfType<ISymbolicDataAnalysisExpressionCrossover<T>>()) {
374        op.EvaluationPartitionParameter.ActualName = FitnessCalculationPartitionParameter.Name;
375        op.ProblemDataParameter.ActualName = ProblemDataParameter.Name;
376        op.EvaluationPartitionParameter.ActualName = FitnessCalculationPartitionParameter.Name;
377        op.RelativeNumberOfEvaluatedSamplesParameter.ActualName = RelativeNumberOfEvaluatedSamplesParameter.Name;
378        op.EvaluatorParameter.ActualName = EvaluatorParameter.Name;
379      }
380    }
381
382    #region Import & Export
383    public virtual void Load(T data) {
384      Name = data.Name;
385      Description = data.Description;
386      ProblemData = data;
387    }
388
389    public virtual T Export() {
390      return ProblemData;
391    }
392    #endregion
393  }
394}
Note: See TracBrowser for help on using the repository browser.