1  #region License Information


2  /* HeuristicLab


3  * Copyright (C) 20022011 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.Generic;


24  using System.Linq;


25  using HeuristicLab.Common;


26  using HeuristicLab.Core;


27  using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;


28  using HeuristicLab.Optimization;


29  using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;


30  using HeuristicLab.Problems.DataAnalysis;


31  using HeuristicLab.Problems.DataAnalysis.Symbolic;


32  using HeuristicLab.Problems.DataAnalysis.Symbolic.Classification;


33 


34  namespace HeuristicLab.Algorithms.DataAnalysis {


35  /// <summary>


36  /// Linear discriminant analysis classification algorithm.


37  /// </summary>


38  [Item("Linear Discriminant Analysis", "Linear discriminant analysis classification algorithm (wrapper for ALGLIB).")]


39  [Creatable("Data Analysis")]


40  [StorableClass]


41  public sealed class LinearDiscriminantAnalysis : FixedDataAnalysisAlgorithm<IClassificationProblem> {


42  private const string LinearDiscriminantAnalysisSolutionResultName = "Linear discriminant analysis solution";


43 


44  [StorableConstructor]


45  private LinearDiscriminantAnalysis(bool deserializing) : base(deserializing) { }


46  private LinearDiscriminantAnalysis(LinearDiscriminantAnalysis original, Cloner cloner)


47  : base(original, cloner) {


48  }


49  public LinearDiscriminantAnalysis()


50  : base() {


51  Problem = new ClassificationProblem();


52  }


53  [StorableHook(HookType.AfterDeserialization)]


54  private void AfterDeserialization() { }


55 


56  public override IDeepCloneable Clone(Cloner cloner) {


57  return new LinearDiscriminantAnalysis(this, cloner);


58  }


59 


60  #region Fisher LDA


61  protected override void Run() {


62  var solution = CreateLinearDiscriminantAnalysisSolution(Problem.ProblemData);


63  Results.Add(new Result(LinearDiscriminantAnalysisSolutionResultName, "The linear discriminant analysis.", solution));


64  }


65 


66  public static IClassificationSolution CreateLinearDiscriminantAnalysisSolution(IClassificationProblemData problemData) {


67  Dataset dataset = problemData.Dataset;


68  string targetVariable = problemData.TargetVariable;


69  IEnumerable<string> allowedInputVariables = problemData.AllowedInputVariables;


70  IEnumerable<int> rows = problemData.TrainingIndizes;


71  int nClasses = problemData.ClassNames.Count();


72  double[,] inputMatrix = AlglibUtil.PrepareInputMatrix(dataset, allowedInputVariables.Concat(new string[] { targetVariable }), rows);


73  if (inputMatrix.Cast<double>().Any(x => double.IsNaN(x)  double.IsInfinity(x)))


74  throw new NotSupportedException("Linear discriminant analysis does not support NaN or infinity values in the input dataset.");


75 


76  // change class values into class index


77  int targetVariableColumn = inputMatrix.GetLength(1)  1;


78  List<double> classValues = problemData.ClassValues.OrderBy(x => x).ToList();


79  for (int row = 0; row < inputMatrix.GetLength(0); row++) {


80  inputMatrix[row, targetVariableColumn] = classValues.IndexOf(inputMatrix[row, targetVariableColumn]);


81  }


82  int info;


83  double[] w;


84  alglib.fisherlda(inputMatrix, inputMatrix.GetLength(0), allowedInputVariables.Count(), nClasses, out info, out w);


85  if (info < 1) throw new ArgumentException("Error in calculation of linear discriminant analysis solution");


86 


87  ISymbolicExpressionTree tree = new SymbolicExpressionTree(new ProgramRootSymbol().CreateTreeNode());


88  ISymbolicExpressionTreeNode startNode = new StartSymbol().CreateTreeNode();


89  tree.Root.AddSubtree(startNode);


90  ISymbolicExpressionTreeNode addition = new Addition().CreateTreeNode();


91  startNode.AddSubtree(addition);


92 


93  int col = 0;


94  foreach (string column in allowedInputVariables) {


95  VariableTreeNode vNode = (VariableTreeNode)new HeuristicLab.Problems.DataAnalysis.Symbolic.Variable().CreateTreeNode();


96  vNode.VariableName = column;


97  vNode.Weight = w[col];


98  addition.AddSubtree(vNode);


99  col++;


100  }


101 


102  var model = LinearDiscriminantAnalysis.CreateDiscriminantFunctionModel(tree, new SymbolicDataAnalysisExpressionTreeInterpreter(), problemData, rows);


103  SymbolicDiscriminantFunctionClassificationSolution solution = new SymbolicDiscriminantFunctionClassificationSolution(model, (IClassificationProblemData)problemData.Clone());


104 


105  return solution;


106  }


107  #endregion


108 


109  private static SymbolicDiscriminantFunctionClassificationModel CreateDiscriminantFunctionModel(ISymbolicExpressionTree tree,


110  ISymbolicDataAnalysisExpressionTreeInterpreter interpreter,


111  IClassificationProblemData problemData,


112  IEnumerable<int> rows) {


113  return new SymbolicDiscriminantFunctionClassificationModel(tree, interpreter);


114  }


115  }


116  }

