[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.Text;
|
---|
| 26 | using System.Threading;
|
---|
| 27 | using HeuristicLab.Common;
|
---|
| 28 | using HeuristicLab.Core;
|
---|
| 29 | using HeuristicLab.Optimization;
|
---|
| 30 | using HeuristicLab.Problems.DataAnalysis;
|
---|
[16847] | 31 | using HEAL.Attic;
|
---|
[15830] | 32 |
|
---|
| 33 | namespace HeuristicLab.Algorithms.DataAnalysis {
|
---|
[16847] | 34 | [StorableType("425AF262-A756-4E9A-B76F-4D2480BEA4FD")]
|
---|
[17081] | 35 | public class RegressionRuleModel : RegressionModel, IDecisionTreeModel {
|
---|
[15830] | 36 | #region Properties
|
---|
| 37 | [Storable]
|
---|
| 38 | public string[] SplitAttributes { get; set; }
|
---|
| 39 | [Storable]
|
---|
| 40 | private double[] SplitValues { get; set; }
|
---|
| 41 | [Storable]
|
---|
| 42 | private Comparison[] Comparisons { get; set; }
|
---|
| 43 | [Storable]
|
---|
| 44 | private IRegressionModel RuleModel { get; set; }
|
---|
| 45 | [Storable]
|
---|
| 46 | private IReadOnlyList<string> variables;
|
---|
| 47 | #endregion
|
---|
| 48 |
|
---|
| 49 | #region HLConstructors
|
---|
| 50 | [StorableConstructor]
|
---|
[16847] | 51 | protected RegressionRuleModel(StorableConstructorFlag _) : base(_) { }
|
---|
[15830] | 52 | protected RegressionRuleModel(RegressionRuleModel original, Cloner cloner) : base(original, cloner) {
|
---|
| 53 | if (original.SplitAttributes != null) SplitAttributes = original.SplitAttributes.ToArray();
|
---|
| 54 | if (original.SplitValues != null) SplitValues = original.SplitValues.ToArray();
|
---|
| 55 | if (original.Comparisons != null) Comparisons = original.Comparisons.ToArray();
|
---|
| 56 | RuleModel = cloner.Clone(original.RuleModel);
|
---|
| 57 | if (original.variables != null) variables = original.variables.ToList();
|
---|
| 58 | }
|
---|
| 59 | private RegressionRuleModel(string target) : base(target) { }
|
---|
| 60 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 61 | return new RegressionRuleModel(this, cloner);
|
---|
| 62 | }
|
---|
| 63 | #endregion
|
---|
| 64 |
|
---|
| 65 | internal static RegressionRuleModel CreateRuleModel(string target, RegressionTreeParameters regressionTreeParams) {
|
---|
| 66 | return regressionTreeParams.LeafModel.ProvidesConfidence ? new ConfidenceRegressionRuleModel(target) : new RegressionRuleModel(target);
|
---|
| 67 | }
|
---|
| 68 |
|
---|
| 69 | #region IRegressionModel
|
---|
| 70 | public override IEnumerable<string> VariablesUsedForPrediction {
|
---|
| 71 | get { return variables; }
|
---|
| 72 | }
|
---|
| 73 |
|
---|
| 74 | public override IEnumerable<double> GetEstimatedValues(IDataset dataset, IEnumerable<int> rows) {
|
---|
| 75 | if (RuleModel == null) throw new NotSupportedException("The model has not been built correctly");
|
---|
| 76 | return RuleModel.GetEstimatedValues(dataset, rows);
|
---|
| 77 | }
|
---|
| 78 |
|
---|
| 79 | public override IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
|
---|
| 80 | return new RegressionSolution(this, problemData);
|
---|
| 81 | }
|
---|
| 82 | #endregion
|
---|
| 83 |
|
---|
| 84 | public void Build(IReadOnlyList<int> trainingRows, IReadOnlyList<int> pruningRows, IScope statescope, ResultCollection results, CancellationToken cancellationToken) {
|
---|
[17080] | 85 | var regressionTreeParams = (RegressionTreeParameters)statescope.Variables[DecisionTreeRegression.RegressionTreeParameterVariableName].Value;
|
---|
[15830] | 86 | variables = regressionTreeParams.AllowedInputVariables.ToList();
|
---|
| 87 |
|
---|
| 88 | //build tree and select node with maximum coverage
|
---|
| 89 | var tree = RegressionNodeTreeModel.CreateTreeModel(regressionTreeParams.TargetVariable, regressionTreeParams);
|
---|
[16847] | 90 | tree.BuildModel(trainingRows, pruningRows, statescope, results, cancellationToken);
|
---|
[15830] | 91 | var nodeModel = tree.Root.EnumerateNodes().Where(x => x.IsLeaf).MaxItems(x => x.NumSamples).First();
|
---|
| 92 |
|
---|
| 93 | var satts = new List<string>();
|
---|
| 94 | var svals = new List<double>();
|
---|
| 95 | var reops = new List<Comparison>();
|
---|
| 96 |
|
---|
[16847] | 97 | //extract splits
|
---|
[15830] | 98 | for (var temp = nodeModel; temp.Parent != null; temp = temp.Parent) {
|
---|
| 99 | satts.Add(temp.Parent.SplitAttribute);
|
---|
| 100 | svals.Add(temp.Parent.SplitValue);
|
---|
| 101 | reops.Add(temp.Parent.Left == temp ? Comparison.LessEqual : Comparison.Greater);
|
---|
| 102 | }
|
---|
| 103 | Comparisons = reops.ToArray();
|
---|
| 104 | SplitAttributes = satts.ToArray();
|
---|
| 105 | SplitValues = svals.ToArray();
|
---|
| 106 | int np;
|
---|
| 107 | RuleModel = regressionTreeParams.LeafModel.BuildModel(trainingRows.Union(pruningRows).Where(r => Covers(regressionTreeParams.Data, r)).ToArray(), regressionTreeParams, cancellationToken, out np);
|
---|
| 108 | }
|
---|
| 109 |
|
---|
| 110 | public void Update(IReadOnlyList<int> rows, IScope statescope, CancellationToken cancellationToken) {
|
---|
[17080] | 111 | var regressionTreeParams = (RegressionTreeParameters)statescope.Variables[DecisionTreeRegression.RegressionTreeParameterVariableName].Value;
|
---|
[15830] | 112 | int np;
|
---|
| 113 | RuleModel = regressionTreeParams.LeafModel.BuildModel(rows, regressionTreeParams, cancellationToken, out np);
|
---|
| 114 | }
|
---|
| 115 |
|
---|
| 116 | public bool Covers(IDataset dataset, int row) {
|
---|
| 117 | return !SplitAttributes.Where((t, i) => !Comparisons[i].Compare(dataset.GetDoubleValue(t, row), SplitValues[i])).Any();
|
---|
| 118 | }
|
---|
| 119 |
|
---|
| 120 | public string ToCompactString() {
|
---|
| 121 | var mins = new Dictionary<string, double>();
|
---|
| 122 | var maxs = new Dictionary<string, double>();
|
---|
| 123 | for (var i = 0; i < SplitAttributes.Length; i++) {
|
---|
| 124 | var n = SplitAttributes[i];
|
---|
| 125 | var v = SplitValues[i];
|
---|
| 126 | if (!mins.ContainsKey(n)) mins.Add(n, double.NegativeInfinity);
|
---|
| 127 | if (!maxs.ContainsKey(n)) maxs.Add(n, double.PositiveInfinity);
|
---|
| 128 | if (Comparisons[i] == Comparison.LessEqual) maxs[n] = Math.Min(maxs[n], v);
|
---|
| 129 | else mins[n] = Math.Max(mins[n], v);
|
---|
| 130 | }
|
---|
| 131 | if (maxs.Count == 0) return "";
|
---|
| 132 | var s = new StringBuilder();
|
---|
| 133 | foreach (var key in maxs.Keys)
|
---|
| 134 | s.Append(string.Format("{0} ∈ [{1:e2}; {2:e2}] && ", key, mins[key], maxs[key]));
|
---|
| 135 | s.Remove(s.Length - 4, 4);
|
---|
| 136 | return s.ToString();
|
---|
| 137 | }
|
---|
| 138 |
|
---|
[16847] | 139 | [StorableType("7302AA30-9F58-42F3-BF6A-ECF1536508AB")]
|
---|
[15830] | 140 | private sealed class ConfidenceRegressionRuleModel : RegressionRuleModel, IConfidenceRegressionModel {
|
---|
| 141 | #region HLConstructors
|
---|
| 142 | [StorableConstructor]
|
---|
[16847] | 143 | private ConfidenceRegressionRuleModel(StorableConstructorFlag _) : base(_) { }
|
---|
[15830] | 144 | private ConfidenceRegressionRuleModel(ConfidenceRegressionRuleModel original, Cloner cloner) : base(original, cloner) { }
|
---|
| 145 | public ConfidenceRegressionRuleModel(string targetAttr) : base(targetAttr) { }
|
---|
| 146 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 147 | return new ConfidenceRegressionRuleModel(this, cloner);
|
---|
| 148 | }
|
---|
| 149 | #endregion
|
---|
| 150 |
|
---|
| 151 | public IEnumerable<double> GetEstimatedVariances(IDataset dataset, IEnumerable<int> rows) {
|
---|
| 152 | return ((IConfidenceRegressionModel)RuleModel).GetEstimatedVariances(dataset, rows);
|
---|
| 153 | }
|
---|
| 154 |
|
---|
| 155 | public override IRegressionSolution CreateRegressionSolution(IRegressionProblemData problemData) {
|
---|
| 156 | return new ConfidenceRegressionSolution(this, problemData);
|
---|
| 157 | }
|
---|
| 158 | }
|
---|
| 159 | }
|
---|
| 160 |
|
---|
[16847] | 161 | [StorableType("152DECE4-2692-4D53-B290-974806ADCD72")]
|
---|
[15830] | 162 | internal enum Comparison {
|
---|
| 163 | LessEqual,
|
---|
| 164 | Greater
|
---|
| 165 | }
|
---|
| 166 |
|
---|
| 167 | internal static class ComparisonExtentions {
|
---|
| 168 | public static bool Compare(this Comparison op, double x, double y) {
|
---|
| 169 | switch (op) {
|
---|
| 170 | case Comparison.Greater:
|
---|
| 171 | return x > y;
|
---|
| 172 | case Comparison.LessEqual:
|
---|
| 173 | return x <= y;
|
---|
| 174 | default:
|
---|
| 175 | throw new ArgumentOutOfRangeException(op.ToString(), op, null);
|
---|
| 176 | }
|
---|
| 177 | }
|
---|
| 178 | }
|
---|
| 179 | } |
---|