[15830] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
[17209] | 3 | * Copyright (C) Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[15830] | 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 System.Linq;
|
---|
| 25 | using System.Threading;
|
---|
| 26 | using HeuristicLab.Common;
|
---|
| 27 | using HeuristicLab.Core;
|
---|
| 28 | using HeuristicLab.Data;
|
---|
| 29 | using HeuristicLab.Optimization;
|
---|
| 30 | using HeuristicLab.Problems.DataAnalysis;
|
---|
[16847] | 31 | using HEAL.Attic;
|
---|
[15830] | 32 |
|
---|
| 33 | namespace HeuristicLab.Algorithms.DataAnalysis {
|
---|
[16847] | 34 | [StorableType("7B4D9AE9-0456-4029-80A6-CCB5E33CE356")]
|
---|
[17081] | 35 | public class RegressionRuleSetModel : RegressionModel, IDecisionTreeModel {
|
---|
[15830] | 36 | private const string NumRulesResultName = "Number of rules";
|
---|
| 37 | private const string CoveredInstancesResultName = "Covered instances";
|
---|
| 38 | public const string RuleSetStateVariableName = "RuleSetState";
|
---|
| 39 |
|
---|
| 40 | #region Properties
|
---|
| 41 | [Storable]
|
---|
| 42 | internal List<RegressionRuleModel> Rules { get; private set; }
|
---|
| 43 | #endregion
|
---|
| 44 |
|
---|
| 45 | #region HLConstructors & Cloning
|
---|
| 46 | [StorableConstructor]
|
---|
[16847] | 47 | protected RegressionRuleSetModel(StorableConstructorFlag _) : base(_) { }
|
---|
[15830] | 48 | protected RegressionRuleSetModel(RegressionRuleSetModel original, Cloner cloner) : base(original, cloner) {
|
---|
| 49 | if (original.Rules != null) Rules = original.Rules.Select(cloner.Clone).ToList();
|
---|
| 50 | }
|
---|
| 51 | protected RegressionRuleSetModel(string targetVariable) : base(targetVariable) { }
|
---|
| 52 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 53 | return new RegressionRuleSetModel(this, cloner);
|
---|
| 54 | }
|
---|
| 55 | #endregion
|
---|
| 56 |
|
---|
| 57 | internal static RegressionRuleSetModel CreateRuleModel(string targetAttr, RegressionTreeParameters regressionTreeParams) {
|
---|
| 58 | return regressionTreeParams.LeafModel.ProvidesConfidence ? new ConfidenceRegressionRuleSetModel(targetAttr) : new RegressionRuleSetModel(targetAttr);
|
---|
| 59 | }
|
---|
| 60 |
|
---|
| 61 | #region RegressionModel
|
---|
| 62 | public override IEnumerable<string> VariablesUsedForPrediction {
|
---|
| 63 | get {
|
---|
| 64 | var f = Rules.FirstOrDefault();
|
---|
| 65 | return f != null ? (f.VariablesUsedForPrediction ?? new List<string>()) : new List<string>();
|
---|
| 66 | }
|
---|
| 67 | }
|
---|
| 68 | public override IEnumerable<double> GetEstimatedValues(IDataset dataset, IEnumerable<int> rows) {
|
---|
| 69 | if (Rules == null) throw new NotSupportedException("The model has not been built yet");
|
---|
| 70 | return rows.Select(row => GetEstimatedValue(dataset, row));
|
---|
| 71 | }
|
---|
| 72 | public override IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
|
---|
| 73 | return new RegressionSolution(this, problemData);
|
---|
| 74 | }
|
---|
| 75 | #endregion
|
---|
| 76 |
|
---|
[17081] | 77 | #region IDecisionTreeModel
|
---|
[15830] | 78 | public void Build(IReadOnlyList<int> trainingRows, IReadOnlyList<int> pruningRows, IScope stateScope, ResultCollection results, CancellationToken cancellationToken) {
|
---|
[17080] | 79 | var regressionTreeParams = (RegressionTreeParameters)stateScope.Variables[DecisionTreeRegression.RegressionTreeParameterVariableName].Value;
|
---|
[15830] | 80 | var ruleSetState = (RuleSetState)stateScope.Variables[RuleSetStateVariableName].Value;
|
---|
| 81 |
|
---|
| 82 | if (ruleSetState.Code <= 0) {
|
---|
| 83 | ruleSetState.Rules.Clear();
|
---|
| 84 | ruleSetState.TrainingRows = trainingRows;
|
---|
| 85 | ruleSetState.PruningRows = pruningRows;
|
---|
| 86 | ruleSetState.Code = 1;
|
---|
| 87 | }
|
---|
| 88 |
|
---|
| 89 | do {
|
---|
| 90 | var tempRule = RegressionRuleModel.CreateRuleModel(regressionTreeParams.TargetVariable, regressionTreeParams);
|
---|
| 91 | cancellationToken.ThrowIfCancellationRequested();
|
---|
| 92 |
|
---|
| 93 | if (!results.ContainsKey(NumRulesResultName)) results.Add(new Result(NumRulesResultName, new IntValue(0)));
|
---|
| 94 | if (!results.ContainsKey(CoveredInstancesResultName)) results.Add(new Result(CoveredInstancesResultName, new IntValue(0)));
|
---|
| 95 |
|
---|
| 96 | var t1 = ruleSetState.TrainingRows.Count;
|
---|
| 97 | tempRule.Build(ruleSetState.TrainingRows, ruleSetState.PruningRows, stateScope, results, cancellationToken);
|
---|
| 98 | ruleSetState.TrainingRows = ruleSetState.TrainingRows.Where(i => !tempRule.Covers(regressionTreeParams.Data, i)).ToArray();
|
---|
| 99 | ruleSetState.PruningRows = ruleSetState.PruningRows.Where(i => !tempRule.Covers(regressionTreeParams.Data, i)).ToArray();
|
---|
| 100 | ruleSetState.Rules.Add(tempRule);
|
---|
| 101 | ((IntValue)results[NumRulesResultName].Value).Value++;
|
---|
| 102 | ((IntValue)results[CoveredInstancesResultName].Value).Value += t1 - ruleSetState.TrainingRows.Count;
|
---|
| 103 | }
|
---|
| 104 | while (ruleSetState.TrainingRows.Count > 0);
|
---|
| 105 | Rules = ruleSetState.Rules;
|
---|
| 106 | }
|
---|
| 107 | public void Update(IReadOnlyList<int> rows, IScope stateScope, CancellationToken cancellationToken) {
|
---|
| 108 | foreach (var rule in Rules) rule.Update(rows, stateScope, cancellationToken);
|
---|
| 109 | }
|
---|
| 110 | public static void Initialize(IScope stateScope) {
|
---|
| 111 | stateScope.Variables.Add(new Variable(RuleSetStateVariableName, new RuleSetState()));
|
---|
| 112 | }
|
---|
| 113 | #endregion
|
---|
| 114 |
|
---|
| 115 | #region Helpers
|
---|
| 116 | private double GetEstimatedValue(IDataset dataset, int row) {
|
---|
| 117 | foreach (var rule in Rules) {
|
---|
| 118 | if (rule.Covers(dataset, row))
|
---|
| 119 | return rule.GetEstimatedValues(dataset, row.ToEnumerable()).Single();
|
---|
| 120 | }
|
---|
| 121 | throw new ArgumentException("Instance is not covered by any rule");
|
---|
| 122 | }
|
---|
| 123 | #endregion
|
---|
| 124 |
|
---|
[16847] | 125 | [StorableType("E114F3C9-3C1F-443D-8270-0E10CE12F2A0")]
|
---|
[15830] | 126 | public class RuleSetState : Item {
|
---|
| 127 | [Storable]
|
---|
| 128 | public List<RegressionRuleModel> Rules = new List<RegressionRuleModel>();
|
---|
| 129 | [Storable]
|
---|
| 130 | public IReadOnlyList<int> TrainingRows = new List<int>();
|
---|
| 131 | [Storable]
|
---|
| 132 | public IReadOnlyList<int> PruningRows = new List<int>();
|
---|
| 133 |
|
---|
| 134 | //State.Code values denote the current action (for pausing)
|
---|
| 135 | //0...nothing has been done;
|
---|
| 136 | //1...splitting nodes;
|
---|
| 137 | [Storable]
|
---|
| 138 | public int Code = 0;
|
---|
| 139 |
|
---|
| 140 | #region HLConstructors & Cloning
|
---|
| 141 | [StorableConstructor]
|
---|
[16847] | 142 | protected RuleSetState(StorableConstructorFlag _) : base(_) { }
|
---|
[15830] | 143 | protected RuleSetState(RuleSetState original, Cloner cloner) : base(original, cloner) {
|
---|
| 144 | Rules = original.Rules.Select(cloner.Clone).ToList();
|
---|
| 145 | TrainingRows = original.TrainingRows.ToList();
|
---|
| 146 | PruningRows = original.PruningRows.ToList();
|
---|
| 147 |
|
---|
| 148 | Code = original.Code;
|
---|
| 149 | }
|
---|
| 150 | public RuleSetState() { }
|
---|
| 151 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 152 | return new RuleSetState(this, cloner);
|
---|
| 153 | }
|
---|
| 154 | #endregion
|
---|
| 155 | }
|
---|
| 156 |
|
---|
[16847] | 157 | [StorableType("52E7992B-94CC-4960-AA82-1A399BE735C6")]
|
---|
[15830] | 158 | private sealed class ConfidenceRegressionRuleSetModel : RegressionRuleSetModel, IConfidenceRegressionModel {
|
---|
| 159 | #region HLConstructors & Cloning
|
---|
| 160 | [StorableConstructor]
|
---|
[16847] | 161 | private ConfidenceRegressionRuleSetModel(StorableConstructorFlag _) : base(_) { }
|
---|
[15830] | 162 | private ConfidenceRegressionRuleSetModel(ConfidenceRegressionRuleSetModel original, Cloner cloner) : base(original, cloner) { }
|
---|
| 163 | public ConfidenceRegressionRuleSetModel(string targetVariable) : base(targetVariable) { }
|
---|
| 164 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 165 | return new ConfidenceRegressionRuleSetModel(this, cloner);
|
---|
| 166 | }
|
---|
| 167 | #endregion
|
---|
| 168 |
|
---|
| 169 | #region IConfidenceRegressionModel
|
---|
| 170 | public IEnumerable<double> GetEstimatedVariances(IDataset dataset, IEnumerable<int> rows) {
|
---|
| 171 | if (Rules == null) throw new NotSupportedException("The model has not been built yet");
|
---|
| 172 | return rows.Select(row => GetEstimatedVariance(dataset, row));
|
---|
| 173 | }
|
---|
| 174 | public override IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
|
---|
| 175 | return new ConfidenceRegressionSolution(this, problemData);
|
---|
| 176 | }
|
---|
| 177 | private double GetEstimatedVariance(IDataset dataset, int row) {
|
---|
| 178 | foreach (var rule in Rules) {
|
---|
| 179 | if (rule.Covers(dataset, row)) return ((IConfidenceRegressionModel)rule).GetEstimatedVariances(dataset, row.ToEnumerable()).Single();
|
---|
| 180 | }
|
---|
| 181 | throw new ArgumentException("Instance is not covered by any rule");
|
---|
| 182 | }
|
---|
| 183 | #endregion
|
---|
| 184 | }
|
---|
| 185 | }
|
---|
| 186 | } |
---|