Free cookie consent management tool by TermsFeed Policy Generator

source: branches/PersistenceOverhaul/HeuristicLab.DataPreprocessing/3.4/ProblemDataCreator.cs @ 17204

Last change on this file since 17204 was 13252, checked in by pfleck, 9 years ago

#2486

  • Added the possibility to add rows or columns by middle-click on a row or column header.
  • Added a rename columns button which shows the new RenameColumnsDialog.
  • Fixed a bug where no input variables where checked after exporting to a DataAnalysisProblemData.
File size: 5.2 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2015 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.Problems.DataAnalysis;
27
28namespace HeuristicLab.DataPreprocessing {
29  public class ProblemDataCreator {
30
31    private readonly IPreprocessingContext context;
32
33    private Dataset ExportedDataset {
34      get {
35        return context.Data.ExportToDataset();
36      }
37    }
38
39    private IList<ITransformation> Transformations { get { return context.Data.Transformations; } }
40
41    public ProblemDataCreator(IPreprocessingContext context) {
42      this.context = context;
43    }
44
45    public IDataAnalysisProblemData CreateProblemData() {
46      if (context.Data.Rows == 0 || context.Data.Columns == 0) return null;
47
48      var oldProblemData = context.ProblemData;
49      IDataAnalysisProblemData problemData;
50
51      if (oldProblemData is RegressionProblemData) {
52        problemData = CreateRegressionData((RegressionProblemData)oldProblemData);
53      } else if (oldProblemData is ClassificationProblemData) {
54        problemData = CreateClassificationData((ClassificationProblemData)oldProblemData);
55      } else if (oldProblemData is ClusteringProblemData) {
56        problemData = CreateClusteringData((ClusteringProblemData)oldProblemData);
57      } else {
58        throw new NotImplementedException("The type of the DataAnalysisProblemData is not supported.");
59      }
60
61      SetTrainingAndTestPartition(problemData);
62      // set the input variables to the correct checked state
63      var inputVariables = oldProblemData.InputVariables.ToDictionary(x => x.Value, x => x);
64      foreach (var variable in problemData.InputVariables) {
65        bool isChecked = inputVariables.ContainsKey(variable.Value) && oldProblemData.InputVariables.ItemChecked(inputVariables[variable.Value]);
66        problemData.InputVariables.SetItemCheckedState(variable, isChecked);
67      }
68
69      return problemData;
70    }
71
72    private IDataAnalysisProblemData CreateRegressionData(RegressionProblemData oldProblemData) {
73      var targetVariable = oldProblemData.TargetVariable;
74      if (!context.Data.VariableNames.Contains(targetVariable))
75        targetVariable = context.Data.VariableNames.First();
76      var inputVariables = GetDoubleInputVariables(targetVariable);
77      var newProblemData = new RegressionProblemData(ExportedDataset, inputVariables, targetVariable, Transformations);
78      return newProblemData;
79    }
80
81    private IDataAnalysisProblemData CreateClassificationData(ClassificationProblemData oldProblemData) {
82      var targetVariable = oldProblemData.TargetVariable;
83      if (!context.Data.VariableNames.Contains(targetVariable))
84        targetVariable = context.Data.VariableNames.First();
85      var inputVariables = GetDoubleInputVariables(targetVariable);
86      var newProblemData = new ClassificationProblemData(ExportedDataset, inputVariables, targetVariable, Transformations);
87      newProblemData.PositiveClass = oldProblemData.PositiveClass;
88      return newProblemData;
89    }
90
91    private IDataAnalysisProblemData CreateClusteringData(ClusteringProblemData oldProblemData) {
92      return new ClusteringProblemData(ExportedDataset, GetDoubleInputVariables(String.Empty), Transformations);
93    }
94
95    private void SetTrainingAndTestPartition(IDataAnalysisProblemData problemData) {
96      var ppData = context.Data;
97
98      problemData.TrainingPartition.Start = ppData.TrainingPartition.Start;
99      problemData.TrainingPartition.End = ppData.TrainingPartition.End;
100      problemData.TestPartition.Start = ppData.TestPartition.Start;
101      problemData.TestPartition.End = ppData.TestPartition.End;
102    }
103
104    private IEnumerable<string> GetDoubleInputVariables(string targetVariable) {
105      var variableNames = new List<string>();
106      for (int i = 0; i < context.Data.Columns; ++i) {
107        var variableName = context.Data.GetVariableName(i);
108        if (context.Data.VariableHasType<double>(i)
109          && variableName != targetVariable
110          && IsNotConstantInputVariable(context.Data.GetValues<double>(i))) {
111
112          variableNames.Add(variableName);
113        }
114      }
115      return variableNames;
116    }
117
118    private bool IsNotConstantInputVariable(IList<double> list) {
119      return context.Data.TrainingPartition.End - context.Data.TrainingPartition.Start > 1 || list.Range() > 0;
120    }
121  }
122}
Note: See TracBrowser for help on using the repository browser.