source: trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/Linear/MultinomialLogitModel.cs @ 14185

Last change on this file since 14185 was 14185, checked in by swagner, 3 years ago

#2526: Updated year of copyrights in license headers

File size: 4.7 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2016 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
22using System;
23using System.Collections.Generic;
24using System.Linq;
25using HeuristicLab.Common;
26using HeuristicLab.Core;
27using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
28using HeuristicLab.Problems.DataAnalysis;
29
30namespace HeuristicLab.Algorithms.DataAnalysis {
31  /// <summary>
32  /// Represents a multinomial logit model for classification
33  /// </summary>
34  [StorableClass]
35  [Item("Multinomial Logit Model", "Represents a multinomial logit model for classification.")]
36  public sealed class MultinomialLogitModel : ClassificationModel {
37
38    private alglib.logitmodel logitModel;
39    public alglib.logitmodel Model {
40      get { return logitModel; }
41      set {
42        if (value != logitModel) {
43          if (value == null) throw new ArgumentNullException();
44          logitModel = value;
45          OnChanged(EventArgs.Empty);
46        }
47      }
48    }
49
50    public override IEnumerable<string> VariablesUsedForPrediction {
51      get { return allowedInputVariables; }
52    }
53
54    [Storable]
55    private string[] allowedInputVariables;
56    [Storable]
57    private double[] classValues;
58    [StorableConstructor]
59    private MultinomialLogitModel(bool deserializing)
60      : base(deserializing) {
61      if (deserializing)
62        logitModel = new alglib.logitmodel();
63    }
64    private MultinomialLogitModel(MultinomialLogitModel original, Cloner cloner)
65      : base(original, cloner) {
66      logitModel = new alglib.logitmodel();
67      logitModel.innerobj.w = (double[])original.logitModel.innerobj.w.Clone();
68      allowedInputVariables = (string[])original.allowedInputVariables.Clone();
69      classValues = (double[])original.classValues.Clone();
70    }
71    public MultinomialLogitModel(alglib.logitmodel logitModel, string targetVariable, IEnumerable<string> allowedInputVariables, double[] classValues)
72      : base(targetVariable) {
73      this.name = ItemName;
74      this.description = ItemDescription;
75      this.logitModel = logitModel;
76      this.allowedInputVariables = allowedInputVariables.ToArray();
77      this.classValues = (double[])classValues.Clone();
78    }
79
80    public override IDeepCloneable Clone(Cloner cloner) {
81      return new MultinomialLogitModel(this, cloner);
82    }
83
84    public override IEnumerable<double> GetEstimatedClassValues(IDataset dataset, IEnumerable<int> rows) {
85      double[,] inputData = AlglibUtil.PrepareInputMatrix(dataset, allowedInputVariables, rows);
86
87      int n = inputData.GetLength(0);
88      int columns = inputData.GetLength(1);
89      double[] x = new double[columns];
90      double[] y = new double[classValues.Length];
91
92      for (int row = 0; row < n; row++) {
93        for (int column = 0; column < columns; column++) {
94          x[column] = inputData[row, column];
95        }
96        alglib.mnlprocess(logitModel, x, ref y);
97        // find class for with the largest probability value
98        int maxProbClassIndex = 0;
99        double maxProb = y[0];
100        for (int i = 1; i < y.Length; i++) {
101          if (maxProb < y[i]) {
102            maxProb = y[i];
103            maxProbClassIndex = i;
104          }
105        }
106        yield return classValues[maxProbClassIndex];
107      }
108    }
109
110    public override IClassificationSolution CreateClassificationSolution(IClassificationProblemData problemData) {
111      return new MultinomialLogitClassificationSolution(this, new ClassificationProblemData(problemData));
112    }
113
114    #region events
115    public event EventHandler Changed;
116    private void OnChanged(EventArgs e) {
117      var handlers = Changed;
118      if (handlers != null)
119        handlers(this, e);
120    }
121    #endregion
122
123    #region persistence
124    [Storable]
125    private double[] LogitModelW {
126      get {
127        return logitModel.innerobj.w;
128      }
129      set {
130        logitModel.innerobj.w = value;
131      }
132    }
133    #endregion
134
135  }
136}
Note: See TracBrowser for help on using the repository browser.