Free cookie consent management tool by TermsFeed Policy Generator

source: branches/symbreg-factors-2650/HeuristicLab.Algorithms.DataAnalysis/3.4/Linear/MultinomialLogitClassification.cs @ 14240

Last change on this file since 14240 was 14240, checked in by gkronber, 8 years ago

#2650: added support for categorical variables to LDA and MNL (TODO: OneR )

File size: 5.5 KB
RevLine 
[6567]1#region License Information
2/* HeuristicLab
[14185]3 * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
[6567]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.Data;
28using HeuristicLab.Optimization;
29using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
30using HeuristicLab.Problems.DataAnalysis;
31
32namespace HeuristicLab.Algorithms.DataAnalysis {
33  /// <summary>
34  /// Multinomial logit regression data analysis algorithm.
35  /// </summary>
[13238]36  [Item("Multinomial Logit Classification (MNL)", "Multinomial logit classification data analysis algorithm (wrapper for ALGLIB).")]
[12622]37  [Creatable(CreatableAttribute.Categories.DataAnalysisClassification, Priority = 180)]
[6567]38  [StorableClass]
39  public sealed class MultiNomialLogitClassification : FixedDataAnalysisAlgorithm<IClassificationProblem> {
40    private const string LogitClassificationModelResultName = "Logit classification solution";
41
42    [StorableConstructor]
43    private MultiNomialLogitClassification(bool deserializing) : base(deserializing) { }
44    private MultiNomialLogitClassification(MultiNomialLogitClassification original, Cloner cloner)
45      : base(original, cloner) {
46    }
47    public MultiNomialLogitClassification()
48      : base() {
49      Problem = new ClassificationProblem();
50    }
51    [StorableHook(HookType.AfterDeserialization)]
52    private void AfterDeserialization() { }
53
54    public override IDeepCloneable Clone(Cloner cloner) {
55      return new MultiNomialLogitClassification(this, cloner);
56    }
57
[6633]58    #region logit classification
[6567]59    protected override void Run() {
60      double rmsError, relClassError;
61      var solution = CreateLogitClassificationSolution(Problem.ProblemData, out rmsError, out relClassError);
[6633]62      Results.Add(new Result(LogitClassificationModelResultName, "The logit classification solution.", solution));
63      Results.Add(new Result("Root mean squared error", "The root of the mean of squared errors of the logit regression solution on the training set.", new DoubleValue(rmsError)));
[6567]64      Results.Add(new Result("Relative classification error", "Relative classification error on the training set (percentage of misclassified cases).", new PercentValue(relClassError)));
65    }
66
67    public static IClassificationSolution CreateLogitClassificationSolution(IClassificationProblemData problemData, out double rmsError, out double relClassError) {
[12509]68      var dataset = problemData.Dataset;
[6567]69      string targetVariable = problemData.TargetVariable;
[14240]70      var doubleVariableNames = problemData.AllowedInputVariables.Where(dataset.VariableHasType<double>);
71      var factorVariableNames = problemData.AllowedInputVariables.Where(dataset.VariableHasType<string>);
[8139]72      IEnumerable<int> rows = problemData.TrainingIndices;
[14240]73      double[,] inputMatrix = AlglibUtil.PrepareInputMatrix(dataset, doubleVariableNames.Concat(new string[] { targetVariable }), rows);
74
75      var factorVariableValues = AlglibUtil.GetFactorVariableValues(dataset, factorVariableNames, rows);
76      var factorMatrix = AlglibUtil.PrepareInputMatrix(dataset, factorVariableValues, rows);
77      inputMatrix = factorMatrix.VertCat(inputMatrix);
78
[6567]79      if (inputMatrix.Cast<double>().Any(x => double.IsNaN(x) || double.IsInfinity(x)))
80        throw new NotSupportedException("Multinomial logit classification does not support NaN or infinity values in the input dataset.");
81
[12817]82      alglib.logitmodel lm = new alglib.logitmodel();
83      alglib.mnlreport rep = new alglib.mnlreport();
[6567]84      int nRows = inputMatrix.GetLength(0);
85      int nFeatures = inputMatrix.GetLength(1) - 1;
[6740]86      double[] classValues = dataset.GetDoubleValues(targetVariable).Distinct().OrderBy(x => x).ToArray();
[6567]87      int nClasses = classValues.Count();
88      // map original class values to values [0..nClasses-1]
[8139]89      Dictionary<double, double> classIndices = new Dictionary<double, double>();
[6567]90      for (int i = 0; i < nClasses; i++) {
[8139]91        classIndices[classValues[i]] = i;
[6567]92      }
93      for (int row = 0; row < nRows; row++) {
[8139]94        inputMatrix[row, nFeatures] = classIndices[inputMatrix[row, nFeatures]];
[6567]95      }
96      int info;
97      alglib.mnltrainh(inputMatrix, nRows, nFeatures, nClasses, out info, out lm, out rep);
98      if (info != 1) throw new ArgumentException("Error in calculation of logit classification solution");
99
100      rmsError = alglib.mnlrmserror(lm, inputMatrix, nRows);
101      relClassError = alglib.mnlrelclserror(lm, inputMatrix, nRows);
102
[14240]103      MultinomialLogitClassificationSolution solution = new MultinomialLogitClassificationSolution(new MultinomialLogitModel(lm, targetVariable, doubleVariableNames, factorVariableValues, classValues), (IClassificationProblemData)problemData.Clone());
[6567]104      return solution;
105    }
106    #endregion
107  }
108}
Note: See TracBrowser for help on using the repository browser.