1  #region License Information


2  /* HeuristicLab


3  * Copyright (C) 20022019 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.Threading;


24  using HeuristicLab.Common;


25  using HeuristicLab.Core;


26  using HeuristicLab.Data;


27  using HeuristicLab.Optimization;


28  using HeuristicLab.Parameters;


29  using HEAL.Attic;


30  using HeuristicLab.Problems.DataAnalysis;


31 


32  namespace HeuristicLab.Algorithms.DataAnalysis {


33  /// <summary>


34  /// Nearest neighbour regression data analysis algorithm.


35  /// </summary>


36  [Item("Nearest Neighbour Regression (kNN)", "Nearest neighbour regression data analysis algorithm (wrapper for ALGLIB).")]


37  [Creatable(CreatableAttribute.Categories.DataAnalysisRegression, Priority = 150)]


38  [StorableType("3F940BE04F444F7FA3EEE47423C7F22D")]


39  public sealed class NearestNeighbourRegression : FixedDataAnalysisAlgorithm<IRegressionProblem> {


40  private const string KParameterName = "K";


41  private const string NearestNeighbourRegressionModelResultName = "Nearest neighbour regression solution";


42  private const string WeightsParameterName = "Weights";


43  private const string SelfMatchParameterName = "SelfMatch";


44 


45  #region parameter properties


46  public IFixedValueParameter<IntValue> KParameter {


47  get { return (IFixedValueParameter<IntValue>)Parameters[KParameterName]; }


48  }


49  public IFixedValueParameter<BoolValue> SelfMatchParameter {


50  get { return (IFixedValueParameter<BoolValue>)Parameters[SelfMatchParameterName]; }


51  }


52  public IValueParameter<DoubleArray> WeightsParameter {


53  get { return (IValueParameter<DoubleArray>)Parameters[WeightsParameterName]; }


54  }


55  #endregion


56  #region properties


57  public int K {


58  get { return KParameter.Value.Value; }


59  set {


60  if (value <= 0) throw new ArgumentException("K must be larger than zero.", "K");


61  else KParameter.Value.Value = value;


62  }


63  }


64  public bool SelfMatch {


65  get { return SelfMatchParameter.Value.Value; }


66  set { SelfMatchParameter.Value.Value = value; }


67  }


68  public DoubleArray Weights {


69  get { return WeightsParameter.Value; }


70  set { WeightsParameter.Value = value; }


71  }


72  #endregion


73 


74  [StorableConstructor]


75  private NearestNeighbourRegression(StorableConstructorFlag _) : base(_) { }


76  private NearestNeighbourRegression(NearestNeighbourRegression original, Cloner cloner)


77  : base(original, cloner) {


78  }


79  public NearestNeighbourRegression()


80  : base() {


81  Parameters.Add(new FixedValueParameter<IntValue>(KParameterName, "The number of nearest neighbours to consider for regression.", new IntValue(3)));


82  Parameters.Add(new OptionalValueParameter<DoubleArray>(WeightsParameterName, "Optional: use weights to specify individual scaling values for all features. If not set the weights are calculated automatically (each feature is scaled to unit variance)"));


83  Parameters.Add(new FixedValueParameter<BoolValue>(SelfMatchParameterName, "Should we use equal points for classification?", new BoolValue(false)));


84  Problem = new RegressionProblem();


85  }


86 


87  [StorableHook(HookType.AfterDeserialization)]


88  private void AfterDeserialization() {


89  // BackwardsCompatibility3.3


90  #region Backwards compatible code, remove with 3.4


91  if (!Parameters.ContainsKey(WeightsParameterName)) {


92  Parameters.Add(new OptionalValueParameter<DoubleArray>(WeightsParameterName, "Optional: use weights to specify individual scaling values for all features. If not set the weights are calculated automatically (each feature is scaled to unit variance)"));


93  }


94  if (!Parameters.ContainsKey(SelfMatchParameterName)) {


95  Parameters.Add(new FixedValueParameter<BoolValue>(SelfMatchParameterName, "Should we use equal points for classification?", new BoolValue(false)));


96  }


97  #endregion


98  }


99 


100  public override IDeepCloneable Clone(Cloner cloner) {


101  return new NearestNeighbourRegression(this, cloner);


102  }


103 


104  #region nearest neighbour


105  protected override void Run(CancellationToken cancellationToken) {


106  double[] weights = null;


107  if (Weights != null) weights = Weights.CloneAsArray();


108  var solution = CreateNearestNeighbourRegressionSolution(Problem.ProblemData, K, SelfMatch, weights);


109  Results.Add(new Result(NearestNeighbourRegressionModelResultName, "The nearest neighbour regression solution.", solution));


110  }


111 


112  public static IRegressionSolution CreateNearestNeighbourRegressionSolution(IRegressionProblemData problemData, int k, bool selfMatch = false, double[] weights = null) {


113  var clonedProblemData = (IRegressionProblemData)problemData.Clone();


114  return new NearestNeighbourRegressionSolution(Train(problemData, k, selfMatch, weights), clonedProblemData);


115  }


116 


117  public static INearestNeighbourModel Train(IRegressionProblemData problemData, int k, bool selfMatch = false, double[] weights = null) {


118  return new NearestNeighbourModel(problemData.Dataset,


119  problemData.TrainingIndices,


120  k,


121  selfMatch,


122  problemData.TargetVariable,


123  problemData.AllowedInputVariables,


124  weights);


125  }


126  #endregion


127  }


128  }

