1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2015 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;
|
---|
24 | using System.Collections.Generic;
|
---|
25 | using System.Linq;
|
---|
26 | using HeuristicLab.Analysis;
|
---|
27 | using HeuristicLab.Common;
|
---|
28 | using HeuristicLab.Core;
|
---|
29 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
30 | using HeuristicLab.Optimization;
|
---|
31 | using HeuristicLab.Parameters;
|
---|
32 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
33 | using HeuristicLab.Problems.DataAnalysis;
|
---|
34 | using HeuristicLab.Problems.DataAnalysis.Symbolic;
|
---|
35 | using HeuristicLab.Problems.Instances;
|
---|
36 |
|
---|
37 |
|
---|
38 | namespace HeuristicLab.Problems.GeneticProgramming.GlucosePrediction {
|
---|
39 | [Item("Blood Glucose Forecast", "See MedGEC Workshop at GECCO 2016")]
|
---|
40 | [Creatable(CreatableAttribute.Categories.GeneticProgrammingProblems, Priority = 999)]
|
---|
41 | [StorableClass]
|
---|
42 | public sealed class Problem : SymbolicExpressionTreeProblem, IRegressionProblem, IProblemInstanceConsumer<IRegressionProblemData>, IProblemInstanceExporter<IRegressionProblemData> {
|
---|
43 |
|
---|
44 | #region parameter names
|
---|
45 | private const string ProblemDataParameterName = "ProblemData";
|
---|
46 | #endregion
|
---|
47 |
|
---|
48 | #region Parameter Properties
|
---|
49 | IParameter IDataAnalysisProblem.ProblemDataParameter { get { return ProblemDataParameter; } }
|
---|
50 |
|
---|
51 | public IValueParameter<IRegressionProblemData> ProblemDataParameter {
|
---|
52 | get { return (IValueParameter<IRegressionProblemData>)Parameters[ProblemDataParameterName]; }
|
---|
53 | }
|
---|
54 | #endregion
|
---|
55 |
|
---|
56 | #region Properties
|
---|
57 | public IRegressionProblemData ProblemData {
|
---|
58 | get { return ProblemDataParameter.Value; }
|
---|
59 | set { ProblemDataParameter.Value = value; }
|
---|
60 | }
|
---|
61 | IDataAnalysisProblemData IDataAnalysisProblem.ProblemData { get { return ProblemData; } }
|
---|
62 | #endregion
|
---|
63 |
|
---|
64 | public event EventHandler ProblemDataChanged;
|
---|
65 |
|
---|
66 | public override bool Maximization {
|
---|
67 | get { return true; }
|
---|
68 | }
|
---|
69 |
|
---|
70 | #region item cloning and persistence
|
---|
71 | // persistence
|
---|
72 | [StorableConstructor]
|
---|
73 | private Problem(bool deserializing) : base(deserializing) { }
|
---|
74 | [StorableHook(HookType.AfterDeserialization)]
|
---|
75 | private void AfterDeserialization() {
|
---|
76 | RegisterEventHandlers();
|
---|
77 | }
|
---|
78 |
|
---|
79 | // cloning
|
---|
80 | private Problem(Problem original, Cloner cloner)
|
---|
81 | : base(original, cloner) {
|
---|
82 | RegisterEventHandlers();
|
---|
83 | }
|
---|
84 | public override IDeepCloneable Clone(Cloner cloner) { return new Problem(this, cloner); }
|
---|
85 | #endregion
|
---|
86 |
|
---|
87 | public Problem()
|
---|
88 | : base() {
|
---|
89 | Parameters.Add(new ValueParameter<IRegressionProblemData>(ProblemDataParameterName, "The data for the glucose prediction problem", new RegressionProblemData()));
|
---|
90 |
|
---|
91 | var g = new SimpleSymbolicExpressionGrammar(); // empty grammar is replaced in UpdateGrammar()
|
---|
92 | base.Encoding = new SymbolicExpressionTreeEncoding(g, 100, 17);
|
---|
93 |
|
---|
94 | UpdateGrammar();
|
---|
95 | RegisterEventHandlers();
|
---|
96 | }
|
---|
97 |
|
---|
98 |
|
---|
99 | public override double Evaluate(ISymbolicExpressionTree tree, IRandom random) {
|
---|
100 | var problemData = ProblemData;
|
---|
101 | var target = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices);
|
---|
102 | var allPredicted = Interpreter.Apply(tree.Root.GetSubtree(0).GetSubtree(0), problemData.Dataset, problemData.AllIndices).ToArray();
|
---|
103 | var predicted = problemData.TrainingIndices.Select(r => allPredicted[r]);
|
---|
104 |
|
---|
105 | // var predicted1 = Interpreter.Apply(tree.Root.GetSubtree(0).GetSubtree(0).GetSubtree(1), problemData.Dataset, rows);
|
---|
106 | // var predicted2 = Interpreter.Apply(tree.Root.GetSubtree(0).GetSubtree(0).GetSubtree(2), problemData.Dataset, rows);
|
---|
107 |
|
---|
108 | var pred0_rsq = Rsq(predicted, target);
|
---|
109 | // var pred1_rsq = Rsq(predicted1, target);
|
---|
110 | // var pred2_rsq = Rsq(predicted2, target);
|
---|
111 | return pred0_rsq; // + pred1_rsq + pred2_rsq;
|
---|
112 | }
|
---|
113 |
|
---|
114 | private double Rsq(IEnumerable<double> predicted, IEnumerable<double> target) {
|
---|
115 | // only take predictions for which the target is not NaN
|
---|
116 | var selectedTuples = target.Zip(predicted, Tuple.Create).Where(t => !double.IsNaN(t.Item1)).ToArray();
|
---|
117 | target = selectedTuples.Select(t => t.Item1);
|
---|
118 | predicted = selectedTuples.Select(t => t.Item2);
|
---|
119 |
|
---|
120 | OnlineCalculatorError errorState;
|
---|
121 | var r = OnlinePearsonsRCalculator.Calculate(target, predicted, out errorState);
|
---|
122 | if (errorState != OnlineCalculatorError.None) r = 0;
|
---|
123 | return r * r;
|
---|
124 | }
|
---|
125 |
|
---|
126 | public override void Analyze(ISymbolicExpressionTree[] trees, double[] qualities, ResultCollection results,
|
---|
127 | IRandom random) {
|
---|
128 | base.Analyze(trees, qualities, results, random);
|
---|
129 |
|
---|
130 | if (!results.ContainsKey("Solution")) {
|
---|
131 | results.Add(new Result("Solution", typeof(IRegressionSolution)));
|
---|
132 | }
|
---|
133 | if (!results.ContainsKey("ScaledTree")) {
|
---|
134 | results.Add(new Result("ScaledTree", typeof(ISymbolicExpressionTree)));
|
---|
135 | }
|
---|
136 | // if (!results.ContainsKey("Terms")) {
|
---|
137 | // results.Add(new Result("Terms", typeof(DataTable)));
|
---|
138 | // }
|
---|
139 |
|
---|
140 | var bestTree = trees.First();
|
---|
141 | var bestQuality = qualities.First();
|
---|
142 | for (int i = 1; i < trees.Length; i++) {
|
---|
143 | if (qualities[i] > bestQuality) {
|
---|
144 | bestQuality = qualities[i];
|
---|
145 | bestTree = trees[i];
|
---|
146 | }
|
---|
147 | }
|
---|
148 |
|
---|
149 | bestTree = (ISymbolicExpressionTree)bestTree.Clone();
|
---|
150 | var expressionNode = bestTree.Root.GetSubtree(0).GetSubtree(0);
|
---|
151 | // scale
|
---|
152 |
|
---|
153 | var problemData = ProblemData;
|
---|
154 | var rows = problemData.AllIndices.ToArray();
|
---|
155 | var target = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, rows).ToArray();
|
---|
156 | var predicted =
|
---|
157 | Interpreter.Apply(expressionNode.GetSubtree(0), problemData.Dataset, rows)
|
---|
158 | .ToArray();
|
---|
159 |
|
---|
160 | var filteredPredicted = rows.Where(r => !double.IsNaN(target[r])).Select(r => predicted[r]).ToArray();
|
---|
161 | var filteredTarget = target.Where(t => !double.IsNaN(t)).ToArray();
|
---|
162 | OnlineCalculatorError error;
|
---|
163 | double alpha;
|
---|
164 | double beta;
|
---|
165 | OnlineLinearScalingParameterCalculator.Calculate(filteredPredicted, filteredTarget, out alpha, out beta, out error);
|
---|
166 |
|
---|
167 | var prod = new SimpleSymbol("*", "*", 2, 2).CreateTreeNode();
|
---|
168 | var sum = new SimpleSymbol("+", "+", 2, 2).CreateTreeNode();
|
---|
169 | var constAlpha = (ConstantTreeNode)(new Constant()).CreateTreeNode();
|
---|
170 | constAlpha.Value = alpha;
|
---|
171 | var constBeta = (ConstantTreeNode)(new Constant()).CreateTreeNode();
|
---|
172 | constBeta.Value = beta;
|
---|
173 |
|
---|
174 | var originalTree = expressionNode.GetSubtree(0);
|
---|
175 | expressionNode.RemoveSubtree(0);
|
---|
176 | expressionNode.AddSubtree(sum);
|
---|
177 | sum.AddSubtree(prod);
|
---|
178 | sum.AddSubtree(constAlpha);
|
---|
179 | prod.AddSubtree(originalTree);
|
---|
180 | prod.AddSubtree(constBeta);
|
---|
181 |
|
---|
182 | var model = new Model(bestTree, problemData.TargetVariable, problemData.AllowedInputVariables.ToArray());
|
---|
183 | model.Name = "Scaled Model";
|
---|
184 | model.Description = "Scaled Model";
|
---|
185 | results["Solution"].Value = model.CreateRegressionSolution(problemData);
|
---|
186 | results["ScaledTree"].Value = bestTree;
|
---|
187 |
|
---|
188 | }
|
---|
189 |
|
---|
190 | #region events
|
---|
191 | private void RegisterEventHandlers() {
|
---|
192 | ProblemDataParameter.ValueChanged += new EventHandler(ProblemDataParameter_ValueChanged);
|
---|
193 | if (ProblemDataParameter.Value != null) ProblemDataParameter.Value.Changed += new EventHandler(ProblemData_Changed);
|
---|
194 | }
|
---|
195 |
|
---|
196 | private void ProblemDataParameter_ValueChanged(object sender, EventArgs e) {
|
---|
197 | ProblemDataParameter.Value.Changed += new EventHandler(ProblemData_Changed);
|
---|
198 | OnProblemDataChanged();
|
---|
199 | OnReset();
|
---|
200 | }
|
---|
201 |
|
---|
202 | private void ProblemData_Changed(object sender, EventArgs e) {
|
---|
203 | OnReset();
|
---|
204 | }
|
---|
205 |
|
---|
206 | private void OnProblemDataChanged() {
|
---|
207 | UpdateGrammar();
|
---|
208 |
|
---|
209 | var handler = ProblemDataChanged;
|
---|
210 | if (handler != null) handler(this, EventArgs.Empty);
|
---|
211 | }
|
---|
212 |
|
---|
213 | private void UpdateGrammar() {
|
---|
214 | // whenever ProblemData is changed we create a new grammar with the necessary symbols
|
---|
215 | var g = new Grammar();
|
---|
216 | Encoding.Grammar = g;
|
---|
217 | }
|
---|
218 | #endregion
|
---|
219 |
|
---|
220 | #region Import & Export
|
---|
221 | public void Load(IRegressionProblemData data) {
|
---|
222 | Name = data.Name;
|
---|
223 | Description = data.Description;
|
---|
224 | ProblemData = data;
|
---|
225 | }
|
---|
226 |
|
---|
227 | public IRegressionProblemData Export() {
|
---|
228 | return ProblemData;
|
---|
229 | }
|
---|
230 | #endregion
|
---|
231 | }
|
---|
232 | }
|
---|