source: trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/SymbolicRegressionSingleObjectiveProblem.cs @ 5847

Last change on this file since 5847 was 5847, checked in by mkommend, 11 years ago

#1418: Adapted data analysis classes to new parameter ctors.

File size: 4.9 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.Linq;
23using HeuristicLab.Common;
24using HeuristicLab.Core;
25using HeuristicLab.Parameters;
26using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
27
28namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression {
29  [Item("Symbolic Regression Problem (single objective)", "Represents a single objective symbolic regression problem.")]
30  [StorableClass]
31  [Creatable("Problems")]
32  public class SymbolicRegressionSingleObjectiveProblem : SymbolicDataAnalysisSingleObjectiveProblem<IRegressionProblemData, ISymbolicRegressionSingleObjectiveEvaluator, ISymbolicDataAnalysisSolutionCreator>, IRegressionProblem {
33    private const double PunishmentFactor = 10;
34    private const int InitialMaximumTreeDepth = 8;
35    private const int InitialMaximumTreeLength = 25;
36    private const string EstimationLimitsParameterName = "EstimationLimits";
37    private const string EstimationLimitsParameterDescription = "The limits for the estimated value that can be returned by the symbolic regression model.";
38
39    #region parameter properties
40    public IFixedValueParameter<DoubleLimit> EstimationLimitsParameter {
41      get { return (IFixedValueParameter<DoubleLimit>)Parameters[EstimationLimitsParameterName]; }
42    }
43    #endregion
44    #region properties
45    public DoubleLimit EstimationLimits {
46      get { return EstimationLimitsParameter.Value; }
47    }
48    #endregion
49    [StorableConstructor]
50    protected SymbolicRegressionSingleObjectiveProblem(bool deserializing) : base(deserializing) { }
51    protected SymbolicRegressionSingleObjectiveProblem(SymbolicRegressionSingleObjectiveProblem original, Cloner cloner) : base(original, cloner) { }
52    public override IDeepCloneable Clone(Cloner cloner) { return new SymbolicRegressionSingleObjectiveProblem(this, cloner); }
53
54    public SymbolicRegressionSingleObjectiveProblem()
55      : base(new RegressionProblemData(), new SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator(), new SymbolicDataAnalysisExpressionTreeCreator()) {
56      Parameters.Add(new FixedValueParameter<DoubleLimit>(EstimationLimitsParameterName, EstimationLimitsParameterDescription));
57
58      Maximization.Value = true;
59      MaximumSymbolicExpressionTreeDepth.Value = InitialMaximumTreeDepth;
60      MaximumSymbolicExpressionTreeLength.Value = InitialMaximumTreeLength;
61
62      InitializeOperators();
63      UpdateEstimationLimits();
64    }
65
66    private void InitializeOperators() {
67      Operators.Add(new SymbolicRegressionSingleObjectiveTrainingBestSolutionAnalyzer());
68      Operators.Add(new SymbolicRegressionSingleObjectiveValidationBestSolutionAnalyzer());
69      Operators.Add(new SymbolicRegressionSingleObjectiveOverfittingAnalyzer());
70      ParameterizeOperators();
71    }
72
73    private void UpdateEstimationLimits() {
74      if (ProblemData.TrainingPartition.Start < ProblemData.TrainingPartition.End) {
75        var targetValues = ProblemData.Dataset.GetVariableValues(ProblemData.TargetVariable, ProblemData.TrainingPartition.Start, ProblemData.TrainingPartition.End);
76        var mean = targetValues.Average();
77        var range = targetValues.Max() - targetValues.Min();
78        EstimationLimits.Upper = mean + PunishmentFactor * range;
79        EstimationLimits.Lower = mean - PunishmentFactor * range;
80      }
81    }
82
83    protected override void OnProblemDataChanged() {
84      base.OnProblemDataChanged();
85      UpdateEstimationLimits();
86    }
87
88    protected override void ParameterizeOperators() {
89      base.ParameterizeOperators();
90      if (Parameters.ContainsKey(EstimationLimitsParameterName)) {
91        var operators = Parameters.OfType<IValueParameter>().Select(p => p.Value).OfType<IOperator>().Union(Operators);
92        foreach (var op in operators.OfType<ISymbolicDataAnalysisBoundedOperator>()) {
93          op.EstimationLimitsParameter.ActualName = EstimationLimitsParameter.Name;
94        }
95      }
96    }
97
98    public override void ImportProblemDataFromFile(string fileName) {
99      RegressionProblemData problemData = RegressionProblemData.ImportFromFile(fileName);
100      ProblemData = problemData;
101    }
102  }
103}
Note: See TracBrowser for help on using the repository browser.