[5557] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
[15583] | 3 | * Copyright (C) 2002-2018 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[5557] | 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 |
|
---|
[13241] | 22 | using System.Collections.Generic;
|
---|
| 23 | using System.Linq;
|
---|
| 24 | using HeuristicLab.Analysis;
|
---|
[5557] | 25 | using HeuristicLab.Common;
|
---|
| 26 | using HeuristicLab.Core;
|
---|
[13241] | 27 | using HeuristicLab.Data;
|
---|
[5557] | 28 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
[13241] | 29 | using HeuristicLab.Optimization;
|
---|
[5557] | 30 | using HeuristicLab.Parameters;
|
---|
| 31 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
| 32 |
|
---|
| 33 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression {
|
---|
| 34 | /// <summary>
|
---|
| 35 | /// An operator that analyzes the training best symbolic regression solution for multi objective symbolic regression problems.
|
---|
| 36 | /// </summary>
|
---|
| 37 | [Item("SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer", "An operator that analyzes the training best symbolic regression solution for multi objective symbolic regression problems.")]
|
---|
| 38 | [StorableClass]
|
---|
[5685] | 39 | public sealed class SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer : SymbolicDataAnalysisMultiObjectiveTrainingBestSolutionAnalyzer<ISymbolicRegressionSolution>,
|
---|
[5747] | 40 | ISymbolicDataAnalysisInterpreterOperator, ISymbolicDataAnalysisBoundedOperator {
|
---|
[5685] | 41 | private const string ProblemDataParameterName = "ProblemData";
|
---|
| 42 | private const string SymbolicDataAnalysisTreeInterpreterParameterName = "SymbolicDataAnalysisTreeInterpreter";
|
---|
[5770] | 43 | private const string EstimationLimitsParameterName = "EstimationLimits";
|
---|
[13241] | 44 | private const string MaximumSymbolicExpressionTreeLengthParameterName = "MaximumSymbolicExpressionTreeLength";
|
---|
| 45 | private const string ValidationPartitionParameterName = "ValidationPartition";
|
---|
[14718] | 46 | private const string AnalyzeTestErrorParameterName = "Analyze Test Error";
|
---|
[13241] | 47 |
|
---|
[5685] | 48 | #region parameter properties
|
---|
| 49 | public ILookupParameter<IRegressionProblemData> ProblemDataParameter {
|
---|
| 50 | get { return (ILookupParameter<IRegressionProblemData>)Parameters[ProblemDataParameterName]; }
|
---|
| 51 | }
|
---|
| 52 | public ILookupParameter<ISymbolicDataAnalysisExpressionTreeInterpreter> SymbolicDataAnalysisTreeInterpreterParameter {
|
---|
| 53 | get { return (ILookupParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>)Parameters[SymbolicDataAnalysisTreeInterpreterParameterName]; }
|
---|
| 54 | }
|
---|
[5770] | 55 | public IValueLookupParameter<DoubleLimit> EstimationLimitsParameter {
|
---|
| 56 | get { return (IValueLookupParameter<DoubleLimit>)Parameters[EstimationLimitsParameterName]; }
|
---|
[5720] | 57 | }
|
---|
[13241] | 58 | public ILookupParameter<IntValue> MaximumSymbolicExpressionTreeLengthParameter {
|
---|
| 59 | get { return (ILookupParameter<IntValue>)Parameters[MaximumSymbolicExpressionTreeLengthParameterName]; }
|
---|
| 60 | }
|
---|
| 61 |
|
---|
| 62 | public IValueLookupParameter<IntRange> ValidationPartitionParameter {
|
---|
| 63 | get { return (IValueLookupParameter<IntRange>)Parameters[ValidationPartitionParameterName]; }
|
---|
| 64 | }
|
---|
[14718] | 65 |
|
---|
| 66 | public IFixedValueParameter<BoolValue> AnalyzeTestErrorParameter {
|
---|
| 67 | get { return (IFixedValueParameter<BoolValue>)Parameters[AnalyzeTestErrorParameterName]; }
|
---|
| 68 | }
|
---|
[5685] | 69 | #endregion
|
---|
| 70 |
|
---|
[14718] | 71 | public bool AnalyzeTestError {
|
---|
| 72 | get { return AnalyzeTestErrorParameter.Value.Value; }
|
---|
| 73 | set { AnalyzeTestErrorParameter.Value.Value = value; }
|
---|
| 74 | }
|
---|
| 75 |
|
---|
[5557] | 76 | [StorableConstructor]
|
---|
| 77 | private SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer(bool deserializing) : base(deserializing) { }
|
---|
| 78 | private SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer(SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer original, Cloner cloner) : base(original, cloner) { }
|
---|
| 79 | public SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer()
|
---|
| 80 | : base() {
|
---|
[13241] | 81 | Parameters.Add(new LookupParameter<IRegressionProblemData>(ProblemDataParameterName, "The problem data for the symbolic regression solution.") { Hidden = true });
|
---|
| 82 | Parameters.Add(new LookupParameter<ISymbolicDataAnalysisExpressionTreeInterpreter>(SymbolicDataAnalysisTreeInterpreterParameterName, "The symbolic data analysis tree interpreter for the symbolic expression tree.") { Hidden = true });
|
---|
| 83 | Parameters.Add(new ValueLookupParameter<DoubleLimit>(EstimationLimitsParameterName, "The lower and upper limit for the estimated values produced by the symbolic regression model.") { Hidden = true });
|
---|
| 84 | Parameters.Add(new LookupParameter<IntValue>(MaximumSymbolicExpressionTreeLengthParameterName, "Maximal length of the symbolic expression.") { Hidden = true });
|
---|
| 85 | Parameters.Add(new ValueLookupParameter<IntRange>(ValidationPartitionParameterName, "The validation partition."));
|
---|
[14718] | 86 | Parameters.Add(new FixedValueParameter<BoolValue>(AnalyzeTestErrorParameterName, "Flag whether the test error should be displayed in the Pareto-Front", new BoolValue(false)));
|
---|
[5557] | 87 | }
|
---|
[5685] | 88 |
|
---|
[13241] | 89 | [StorableHook(HookType.AfterDeserialization)]
|
---|
| 90 | private void AfterDeserialization() {
|
---|
| 91 | if (!Parameters.ContainsKey(MaximumSymbolicExpressionTreeLengthParameterName))
|
---|
| 92 | Parameters.Add(new LookupParameter<IntValue>(MaximumSymbolicExpressionTreeLengthParameterName, "Maximal length of the symbolic expression.") { Hidden = true });
|
---|
| 93 | if (!Parameters.ContainsKey(ValidationPartitionParameterName))
|
---|
| 94 | Parameters.Add(new ValueLookupParameter<IntRange>(ValidationPartitionParameterName, "The validation partition."));
|
---|
[14718] | 95 | if (!Parameters.ContainsKey(AnalyzeTestErrorParameterName))
|
---|
| 96 | Parameters.Add(new FixedValueParameter<BoolValue>(AnalyzeTestErrorParameterName, "Flag whether the test error should be displayed in the Pareto-Front", new BoolValue(false)));
|
---|
[13241] | 97 | }
|
---|
| 98 |
|
---|
[5557] | 99 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 100 | return new SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer(this, cloner);
|
---|
| 101 | }
|
---|
| 102 |
|
---|
| 103 | protected override ISymbolicRegressionSolution CreateSolution(ISymbolicExpressionTree bestTree, double[] bestQuality) {
|
---|
[13941] | 104 | var model = new SymbolicRegressionModel(ProblemDataParameter.ActualValue.TargetVariable, (ISymbolicExpressionTree)bestTree.Clone(), SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, EstimationLimitsParameter.ActualValue.Lower, EstimationLimitsParameter.ActualValue.Upper);
|
---|
[8972] | 105 | if (ApplyLinearScalingParameter.ActualValue.Value) model.Scale(ProblemDataParameter.ActualValue);
|
---|
[5914] | 106 | return new SymbolicRegressionSolution(model, (IRegressionProblemData)ProblemDataParameter.ActualValue.Clone());
|
---|
[5557] | 107 | }
|
---|
[13241] | 108 |
|
---|
| 109 | public override IOperation Apply() {
|
---|
| 110 | var operation = base.Apply();
|
---|
| 111 | var paretoFront = TrainingBestSolutionsParameter.ActualValue;
|
---|
| 112 |
|
---|
| 113 | IResult result;
|
---|
| 114 | ScatterPlot qualityToTreeSize;
|
---|
| 115 | if (!ResultCollection.TryGetValue("Pareto Front Analysis", out result)) {
|
---|
| 116 | qualityToTreeSize = new ScatterPlot("Quality vs Tree Size", "");
|
---|
| 117 | qualityToTreeSize.VisualProperties.XAxisMinimumAuto = false;
|
---|
| 118 | qualityToTreeSize.VisualProperties.XAxisMaximumAuto = false;
|
---|
| 119 | qualityToTreeSize.VisualProperties.YAxisMinimumAuto = false;
|
---|
| 120 | qualityToTreeSize.VisualProperties.YAxisMaximumAuto = false;
|
---|
| 121 |
|
---|
| 122 | qualityToTreeSize.VisualProperties.XAxisMinimumFixedValue = 0;
|
---|
| 123 | qualityToTreeSize.VisualProperties.XAxisMaximumFixedValue = MaximumSymbolicExpressionTreeLengthParameter.ActualValue.Value;
|
---|
| 124 | qualityToTreeSize.VisualProperties.YAxisMinimumFixedValue = 0;
|
---|
| 125 | qualityToTreeSize.VisualProperties.YAxisMaximumFixedValue = 2;
|
---|
| 126 | ResultCollection.Add(new Result("Pareto Front Analysis", qualityToTreeSize));
|
---|
| 127 | } else {
|
---|
| 128 | qualityToTreeSize = (ScatterPlot)result.Value;
|
---|
| 129 | }
|
---|
| 130 |
|
---|
| 131 |
|
---|
| 132 | int previousTreeLength = -1;
|
---|
| 133 | var sizeParetoFront = new LinkedList<ISymbolicRegressionSolution>();
|
---|
| 134 | foreach (var solution in paretoFront.OrderBy(s => s.Model.SymbolicExpressionTree.Length)) {
|
---|
| 135 | int treeLength = solution.Model.SymbolicExpressionTree.Length;
|
---|
| 136 | if (!sizeParetoFront.Any()) sizeParetoFront.AddLast(solution);
|
---|
| 137 | if (solution.TrainingNormalizedMeanSquaredError < sizeParetoFront.Last.Value.TrainingNormalizedMeanSquaredError) {
|
---|
| 138 | if (treeLength == previousTreeLength)
|
---|
| 139 | sizeParetoFront.RemoveLast();
|
---|
| 140 | sizeParetoFront.AddLast(solution);
|
---|
| 141 | }
|
---|
| 142 | previousTreeLength = treeLength;
|
---|
| 143 | }
|
---|
| 144 |
|
---|
| 145 | qualityToTreeSize.Rows.Clear();
|
---|
[14861] | 146 | var trainingRow = new ScatterPlotDataRow("Training NMSE", "", sizeParetoFront.Select(x => new Point2D<double>(x.Model.SymbolicExpressionTree.Length, x.TrainingNormalizedMeanSquaredError, x)));
|
---|
[13241] | 147 | trainingRow.VisualProperties.PointSize = 8;
|
---|
| 148 | qualityToTreeSize.Rows.Add(trainingRow);
|
---|
| 149 |
|
---|
[14718] | 150 | if (AnalyzeTestError) {
|
---|
| 151 | var testRow = new ScatterPlotDataRow("Test NMSE", "",
|
---|
[14861] | 152 | sizeParetoFront.Select(x => new Point2D<double>(x.Model.SymbolicExpressionTree.Length, x.TestNormalizedMeanSquaredError, x)));
|
---|
[14718] | 153 | testRow.VisualProperties.PointSize = 8;
|
---|
| 154 | qualityToTreeSize.Rows.Add(testRow);
|
---|
| 155 | }
|
---|
| 156 |
|
---|
[13241] | 157 | var validationPartition = ValidationPartitionParameter.ActualValue;
|
---|
| 158 | if (validationPartition.Size != 0) {
|
---|
| 159 | var problemData = ProblemDataParameter.ActualValue;
|
---|
| 160 | var validationIndizes = Enumerable.Range(validationPartition.Start, validationPartition.Size).ToList();
|
---|
| 161 | var targetValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, validationIndizes).ToList();
|
---|
| 162 | OnlineCalculatorError error;
|
---|
| 163 | var validationRow = new ScatterPlotDataRow("Validation NMSE", "",
|
---|
| 164 | sizeParetoFront.Select(x => new Point2D<double>(x.Model.SymbolicExpressionTree.Length,
|
---|
| 165 | OnlineNormalizedMeanSquaredErrorCalculator.Calculate(targetValues, x.GetEstimatedValues(validationIndizes), out error))));
|
---|
| 166 | validationRow.VisualProperties.PointSize = 7;
|
---|
| 167 | qualityToTreeSize.Rows.Add(validationRow);
|
---|
| 168 | }
|
---|
| 169 |
|
---|
| 170 | return operation;
|
---|
| 171 | }
|
---|
| 172 |
|
---|
[5557] | 173 | }
|
---|
| 174 | }
|
---|