Free cookie consent management tool by TermsFeed Policy Generator

source: branches/PersistenceSpeedUp/HeuristicLab.Problems.DataAnalysis.Regression/3.3/SupportVectorRegression/SupportVectorRegressionSolution.cs @ 17958

Last change on this file since 17958 was 5692, checked in by gkronber, 14 years ago

#1426 merged r5690 from data analysis refactoring branch (see #1418) into trunk to fix persistence problems of SVMs.

File size: 5.0 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2011 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.Drawing;
25using System.Linq;
26using HeuristicLab.Common;
27using HeuristicLab.Core;
28using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
29using HeuristicLab.Problems.DataAnalysis.SupportVectorMachine;
30using SVM;
31
32namespace HeuristicLab.Problems.DataAnalysis.Regression.SupportVectorRegression {
33  /// <summary>
34  /// Represents a support vector solution for a regression problem which can be visualized in the GUI.
35  /// </summary>
36  [Item("SupportVectorRegressionSolution", "Represents a support vector solution for a regression problem which can be visualized in the GUI.")]
37  [StorableClass]
38  public sealed class SupportVectorRegressionSolution : DataAnalysisSolution {
39    public override Image ItemImage {
40      get { return HeuristicLab.Common.Resources.VSImageLibrary.Function; }
41    }
42
43    public new SupportVectorMachineModel Model {
44      get { return (SupportVectorMachineModel)base.Model; }
45      set { base.Model = value; }
46    }
47
48    private List<double> estimatedValues;
49    public override IEnumerable<double> EstimatedValues {
50      get {
51        if (estimatedValues == null) RecalculateEstimatedValues();
52        return estimatedValues;
53      }
54    }
55
56    public override IEnumerable<double> EstimatedTrainingValues {
57      get {
58        return GetEstimatedValues(ProblemData.TrainingIndizes);
59      }
60    }
61
62    public override IEnumerable<double> EstimatedTestValues {
63      get {
64        return GetEstimatedValues(ProblemData.TestIndizes);
65      }
66    }
67
68    public Dataset SupportVectors {
69      get {
70        int nCol = inputVariables.Count;
71        double[,] data = new double[Model.Model.SupportVectorCount, nCol];
72        int row = 0;
73        foreach (var sv in Model.SupportVectors) {
74          for (int col = 0; col < nCol; col++) {
75            data[row, col] = sv[col];
76          }
77          row++;
78        }
79        return new Dataset(inputVariables, data);
80      }
81    }
82
83    [Storable]
84    private List<string> inputVariables;
85
86    [StorableConstructor]
87    private SupportVectorRegressionSolution(bool deserializing) : base(deserializing) { }
88    private SupportVectorRegressionSolution(SupportVectorRegressionSolution original, Cloner cloner)
89      : base(original, cloner) {
90      this.inputVariables = new List<string>(original.inputVariables);
91    }
92    public SupportVectorRegressionSolution() : base() { }
93    public SupportVectorRegressionSolution(DataAnalysisProblemData problemData, SupportVectorMachineModel model, IEnumerable<string> inputVariables, double lowerEstimationLimit, double upperEstimationLimit)
94      : base(problemData, lowerEstimationLimit, upperEstimationLimit) {
95      this.Model = model;
96      this.inputVariables = new List<string>(inputVariables);
97    }
98
99    [StorableHook(HookType.AfterDeserialization)]
100    private void AfterDeserialization() {
101      #region backwards compatibility
102      if (inputVariables == null) inputVariables = ProblemData.InputVariables.CheckedItems.Select(x => x.Value.Value).ToList();
103      #endregion
104    }
105
106    public override IDeepCloneable Clone(Cloner cloner) {
107      return new SupportVectorRegressionSolution(this, cloner);
108    }
109
110    protected override void RecalculateEstimatedValues() {
111      SVM.Problem problem = SupportVectorMachineUtil.CreateSvmProblem(ProblemData, Enumerable.Range(0, ProblemData.Dataset.Rows));
112      SVM.Problem scaledProblem = Scaling.Scale(Model.RangeTransform, problem);
113
114      estimatedValues = (from row in Enumerable.Range(0, scaledProblem.Count)
115                         let prediction = SVM.Prediction.Predict(Model.Model, scaledProblem.X[row])
116                         let boundedX = Math.Min(UpperEstimationLimit, Math.Max(LowerEstimationLimit, prediction))
117                         select double.IsNaN(boundedX) ? UpperEstimationLimit : boundedX).ToList();
118      OnEstimatedValuesChanged();
119    }
120
121
122    private IEnumerable<double> GetEstimatedValues(IEnumerable<int> rows) {
123      if (estimatedValues == null) RecalculateEstimatedValues();
124      foreach (int row in rows)
125        yield return estimatedValues[row];
126    }
127  }
128}
Note: See TracBrowser for help on using the repository browser.