Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/sources/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Classification/CSV/ClassifiactionCSVInstanceProvider.cs @ 8599

Last change on this file since 8599 was 8599, checked in by sforsten, 12 years ago

#1942: Training and test partition can be defined (with a TrackBar in percent), when importing a csv file for data analysis problems.

File size: 5.7 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2012 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;
24using System.Collections.Generic;
25using System.Globalization;
26using System.IO;
27using System.Linq;
28using System.Text;
29using HeuristicLab.Common;
30using HeuristicLab.Problems.DataAnalysis;
31
32namespace HeuristicLab.Problems.Instances.DataAnalysis {
33  public class ClassificationCSVInstanceProvider : ClassificationInstanceProvider {
34    public override string Name {
35      get { return "CSV File"; }
36    }
37    public override string Description {
38      get {
39        return "";
40      }
41    }
42    public override Uri WebLink {
43      get { return new Uri("http://dev.heuristiclab.com/trac/hl/core/wiki/UsersFAQ#DataAnalysisImportFileFormat"); }
44    }
45    public override string ReferencePublication {
46      get { return ""; }
47    }
48
49    public override IEnumerable<IDataDescriptor> GetDataDescriptors() {
50      return new List<IDataDescriptor>();
51    }
52
53    public override IClassificationProblemData LoadData(IDataDescriptor descriptor) {
54      throw new NotImplementedException();
55    }
56
57    public override bool CanImportData {
58      get { return true; }
59    }
60    public override IClassificationProblemData ImportData(string path) {
61      TableFileParser csvFileParser = new TableFileParser();
62
63      csvFileParser.Parse(path);
64
65      Dataset dataset = new Dataset(csvFileParser.VariableNames, csvFileParser.Values);
66      string targetVar = dataset.DoubleVariables.Last();
67
68      // turn of input variables that are constant in the training partition
69      var allowedInputVars = new List<string>();
70      var trainingIndizes = Enumerable.Range(0, (csvFileParser.Rows * 2) / 3);
71      foreach (var variableName in dataset.DoubleVariables) {
72        if (trainingIndizes.Count() >= 2 && dataset.GetDoubleValues(variableName, trainingIndizes).Range() > 0 &&
73          variableName != targetVar)
74          allowedInputVars.Add(variableName);
75      }
76
77      ClassificationProblemData classificationData = new ClassificationProblemData(dataset, allowedInputVars, targetVar);
78
79      int trainingPartEnd = trainingIndizes.Last();
80      classificationData.TrainingPartition.Start = trainingIndizes.First();
81      classificationData.TrainingPartition.End = trainingPartEnd;
82      classificationData.TestPartition.Start = trainingPartEnd;
83      classificationData.TestPartition.End = csvFileParser.Rows;
84
85      classificationData.Name = Path.GetFileName(path);
86
87      return classificationData;
88    }
89
90    public override IClassificationProblemData ImportData(string path, DataAnalysisImportType type) {
91      TableFileParser csvFileParser = new TableFileParser();
92      csvFileParser.Parse(path);
93
94      List<IList> values = csvFileParser.Values;
95      if (type.Shuffle) {
96        values = Shuffle(values);
97      }
98
99      Dataset dataset = new Dataset(csvFileParser.VariableNames, values);
100      string targetVar = dataset.DoubleVariables.Last();
101
102      // turn of input variables that are constant in the training partition
103      var allowedInputVars = new List<string>();
104      int trainingPartEnd = (csvFileParser.Rows * type.Training) / 100;
105      var trainingIndizes = Enumerable.Range(0, trainingPartEnd);
106      foreach (var variableName in dataset.DoubleVariables) {
107        if (trainingIndizes.Count() >= 2 && dataset.GetDoubleValues(variableName, trainingIndizes).Range() > 0 &&
108          variableName != targetVar)
109          allowedInputVars.Add(variableName);
110      }
111
112      ClassificationProblemData classificationData = new ClassificationProblemData(dataset, allowedInputVars, targetVar);
113
114      classificationData.TrainingPartition.Start = 0;
115      classificationData.TrainingPartition.End = trainingPartEnd;
116      classificationData.TestPartition.Start = trainingPartEnd;
117      classificationData.TestPartition.End = csvFileParser.Rows;
118
119      classificationData.Name = Path.GetFileName(path);
120
121      return classificationData;
122    }
123
124    public override bool CanExportData {
125      get { return true; }
126    }
127    public override void ExportData(IClassificationProblemData instance, string path) {
128      var strBuilder = new StringBuilder();
129
130      foreach (var variable in instance.InputVariables) {
131        strBuilder.Append(variable + CultureInfo.CurrentCulture.TextInfo.ListSeparator);
132      }
133      strBuilder.Remove(strBuilder.Length - CultureInfo.CurrentCulture.TextInfo.ListSeparator.Length, CultureInfo.CurrentCulture.TextInfo.ListSeparator.Length);
134      strBuilder.AppendLine();
135
136      var dataset = instance.Dataset;
137
138      for (int i = 0; i < dataset.Rows; i++) {
139        for (int j = 0; j < dataset.Columns; j++) {
140          if (j > 0) strBuilder.Append(CultureInfo.CurrentCulture.TextInfo.ListSeparator);
141          strBuilder.Append(dataset.GetValue(i, j));
142        }
143        strBuilder.AppendLine();
144      }
145
146      using (var writer = new StreamWriter(path)) {
147        writer.Write(strBuilder);
148      }
149    }
150  }
151}
Note: See TracBrowser for help on using the repository browser.