1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2018 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 |
|
---|
22 | using System;
|
---|
23 | using System.Collections.Generic;
|
---|
24 | using HeuristicLab.Common;
|
---|
25 | using HeuristicLab.Core;
|
---|
26 | using HeuristicLab.Data;
|
---|
27 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
28 | using HEAL.Attic;
|
---|
29 |
|
---|
30 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression {
|
---|
31 | [Item("Pearson R² Constraint Evaluator", "Calculates the square of the pearson correlation coefficient (also known as coefficient of determination) of a symbolic regression solution.")]
|
---|
32 | [StorableType("D61462E4-2032-4790-B63D-5E6512987F64")]
|
---|
33 | public class SymbolicRegressionSingleObjectiveConstraintPearsonRSquaredEvaluator : SymbolicRegressionSingleObjectiveEvaluator {
|
---|
34 | [StorableConstructor]
|
---|
35 | protected SymbolicRegressionSingleObjectiveConstraintPearsonRSquaredEvaluator(StorableConstructorFlag _) : base(_) { }
|
---|
36 | protected SymbolicRegressionSingleObjectiveConstraintPearsonRSquaredEvaluator(SymbolicRegressionSingleObjectiveConstraintPearsonRSquaredEvaluator original, Cloner cloner)
|
---|
37 | : base(original, cloner) {
|
---|
38 | }
|
---|
39 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
40 | return new SymbolicRegressionSingleObjectiveConstraintPearsonRSquaredEvaluator(this, cloner);
|
---|
41 | }
|
---|
42 |
|
---|
43 | public SymbolicRegressionSingleObjectiveConstraintPearsonRSquaredEvaluator() : base() { }
|
---|
44 |
|
---|
45 | public override bool Maximization { get { return true; } }
|
---|
46 |
|
---|
47 | public override IOperation InstrumentedApply() {
|
---|
48 | var solution = SymbolicExpressionTreeParameter.ActualValue;
|
---|
49 | IEnumerable<int> rows = GenerateRowsToEvaluate();
|
---|
50 |
|
---|
51 | double quality = Calculate(SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, solution, EstimationLimitsParameter.ActualValue.Lower, EstimationLimitsParameter.ActualValue.Upper, ProblemDataParameter.ActualValue, rows, ApplyLinearScalingParameter.ActualValue.Value);
|
---|
52 | QualityParameter.ActualValue = new DoubleValue(quality);
|
---|
53 | return base.InstrumentedApply();
|
---|
54 | }
|
---|
55 |
|
---|
56 | public static double Calculate(ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, ISymbolicExpressionTree solution, double lowerEstimationLimit, double upperEstimationLimit, IRegressionProblemData problemData, IEnumerable<int> rows, bool applyLinearScaling) {
|
---|
57 | IEnumerable<double> estimatedValues = interpreter.GetSymbolicExpressionTreeValues(solution, problemData.Dataset, rows);
|
---|
58 | IEnumerable<double> targetValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, rows);
|
---|
59 | OnlineCalculatorError errorState = OnlineCalculatorError.None;
|
---|
60 |
|
---|
61 | IntervalConstraintsParser parser = new IntervalConstraintsParser();
|
---|
62 | var constraints = parser.Parse(((RegressionProblemData) problemData).IntervalConstraintsParameter.Value.Value);
|
---|
63 | var intervalInterpreter = new IntervalInterpreter();
|
---|
64 | var variableRanges = ((RegressionProblemData) problemData).VariableRangesParameter.Value.VariableIntervals;
|
---|
65 |
|
---|
66 |
|
---|
67 | //foreach (var constraint in constraints) {
|
---|
68 | // if (constraint.Variable != null && !variableRanges.ContainsKey(constraint.Variable))
|
---|
69 | // throw new ArgumentException($"The given variable {constraint.Variable} in the constraint does not exists in the model.", nameof(IntervalConstraintsParser));
|
---|
70 | // if (!constraint.IsDerivation) {
|
---|
71 | // var res = intervalInterpreter.GetSymbolicExressionTreeInterval(solution, variableRanges);
|
---|
72 | // if (!constraint.Interval.Contains(res, constraint.InclusiveLowerBound,
|
---|
73 | // constraint.InclusiveUpperBound)) {
|
---|
74 | // return 0;
|
---|
75 | // }
|
---|
76 | // } else {
|
---|
77 | // var tree = solution;
|
---|
78 | // for (var i = 0; i < constraint.NumberOfDerivation; ++i) {
|
---|
79 | // tree = DerivativeCalculator.Derive(tree, constraint.Variable);
|
---|
80 | // }
|
---|
81 | // var res = intervalInterpreter.GetSymbolicExressionTreeInterval(tree, variableRanges);
|
---|
82 | // if (!constraint.Interval.Contains(res, constraint.InclusiveLowerBound,
|
---|
83 | // constraint.InclusiveUpperBound)) {
|
---|
84 | // return 0;
|
---|
85 | // }
|
---|
86 | // }
|
---|
87 | //}
|
---|
88 | // TODO
|
---|
89 | // m = new SymbolicRegressionModel(...)
|
---|
90 | // m.Scale();
|
---|
91 |
|
---|
92 | // var e = m.GetEstimatedValues (TRAINING)
|
---|
93 | // OnlinePearsonCalc.Calculate(e, TARGET_TRAIING)
|
---|
94 |
|
---|
95 | // scaledTree = model.Tree;
|
---|
96 |
|
---|
97 | // constraints mit scaledTree berechnen (auch die Ableitungen)
|
---|
98 |
|
---|
99 | double r;
|
---|
100 | if (applyLinearScaling) {
|
---|
101 | var rCalculator = new OnlinePearsonsRCalculator();
|
---|
102 | CalculateWithScaling(targetValues, estimatedValues, lowerEstimationLimit, upperEstimationLimit, rCalculator, problemData.Dataset.Rows);
|
---|
103 | var model = new SymbolicRegressionModel(problemData.TargetVariable, solution, interpreter, lowerEstimationLimit, upperEstimationLimit);
|
---|
104 | model.Scale(problemData);
|
---|
105 | var e = model.GetEstimatedValues(problemData.Dataset, problemData.TrainingIndices);
|
---|
106 | var scaledTree = model.SymbolicExpressionTree;
|
---|
107 | errorState = rCalculator.ErrorState;
|
---|
108 | if (CheckConstraintsViolations(constraints, intervalInterpreter, variableRanges, scaledTree, out var val)) {
|
---|
109 | r = val;
|
---|
110 | } else {
|
---|
111 | r = rCalculator.R;
|
---|
112 | }
|
---|
113 | } else {
|
---|
114 | IEnumerable<double> boundedEstimatedValues = estimatedValues.LimitToRange(lowerEstimationLimit, upperEstimationLimit);
|
---|
115 | if (CheckConstraintsViolations(constraints, intervalInterpreter, variableRanges, solution, out var val)) {
|
---|
116 | r = val;
|
---|
117 | } else {
|
---|
118 | r = OnlinePearsonsRCalculator.Calculate(targetValues, boundedEstimatedValues, out errorState);
|
---|
119 | }
|
---|
120 | }
|
---|
121 | if (errorState != OnlineCalculatorError.None) return double.NaN;
|
---|
122 | return r*r;
|
---|
123 | }
|
---|
124 |
|
---|
125 |
|
---|
126 | public override double Evaluate(IExecutionContext context, ISymbolicExpressionTree tree,
|
---|
127 | IRegressionProblemData problemData, IEnumerable<int> rows) {
|
---|
128 | SymbolicDataAnalysisTreeInterpreterParameter.ExecutionContext = context;
|
---|
129 | EstimationLimitsParameter.ExecutionContext = context;
|
---|
130 | ApplyLinearScalingParameter.ExecutionContext = context;
|
---|
131 |
|
---|
132 | double r2 = Calculate(SymbolicDataAnalysisTreeInterpreterParameter.ActualValue, tree,
|
---|
133 | EstimationLimitsParameter.ActualValue.Lower, EstimationLimitsParameter.ActualValue.Upper, problemData, rows,
|
---|
134 | ApplyLinearScalingParameter.ActualValue.Value);
|
---|
135 |
|
---|
136 | SymbolicDataAnalysisTreeInterpreterParameter.ExecutionContext = null;
|
---|
137 | EstimationLimitsParameter.ExecutionContext = null;
|
---|
138 | ApplyLinearScalingParameter.ExecutionContext = null;
|
---|
139 |
|
---|
140 | return r2;
|
---|
141 | }
|
---|
142 |
|
---|
143 | private static bool CheckConstraintsViolations(List<IntervalConstraint> constraints, IntervalInterpreter intervalInterpreter,
|
---|
144 | Dictionary<string, Interval> variableRanges, ISymbolicExpressionTree solution, out double r) {
|
---|
145 | foreach (var constraint in constraints) {
|
---|
146 | if (constraint.Variable != null && !variableRanges.ContainsKey(constraint.Variable))
|
---|
147 | throw new ArgumentException($"The given variable {constraint.Variable} in the constraint does not exists in the model.", nameof(IntervalConstraintsParser));
|
---|
148 | if (!constraint.IsDerivation) {
|
---|
149 | var res = intervalInterpreter.GetSymbolicExressionTreeInterval(solution, variableRanges);
|
---|
150 | if (!constraint.Interval.Contains(res, constraint.InclusiveLowerBound,
|
---|
151 | constraint.InclusiveUpperBound)) {
|
---|
152 | r = 0;
|
---|
153 | return true;
|
---|
154 | }
|
---|
155 | } else {
|
---|
156 | var tree = solution;
|
---|
157 | for (var i = 0; i < constraint.NumberOfDerivation; ++i) {
|
---|
158 | tree = DerivativeCalculator.Derive(tree, constraint.Variable);
|
---|
159 | }
|
---|
160 | var res = intervalInterpreter.GetSymbolicExressionTreeInterval(tree, variableRanges);
|
---|
161 | if (!constraint.Interval.Contains(res, constraint.InclusiveLowerBound,
|
---|
162 | constraint.InclusiveUpperBound)) {
|
---|
163 | r = 0;
|
---|
164 | return true;
|
---|
165 | }
|
---|
166 | }
|
---|
167 | }
|
---|
168 | r = 1;
|
---|
169 | return false;
|
---|
170 | }
|
---|
171 | }
|
---|
172 | }
|
---|