source: branches/DataPreprocessing/HeuristicLab.DataPreprocessing/3.4/PreprocessingTransformator.cs @ 11068

Last change on this file since 11068 was 11068, checked in by mkommend, 8 years ago

#2206: Removed seperate plugin for transformations and integrated them in HeuristicLab.Problems.DataAnalysis.

File size: 6.9 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2013 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 System.Text;
26using HeuristicLab.Data;
27using HeuristicLab.Problems.DataAnalysis;
28
29namespace HeuristicLab.DataPreprocessing {
30  public class PreprocessingTransformator {
31    private readonly ITransactionalPreprocessingData preprocessingData;
32
33    private readonly IDictionary<string, IList<double>> originalColumns;
34
35    private readonly IDictionary<string, string> renamedColumns;
36
37    public PreprocessingTransformator(IPreprocessingData preprocessingData) {
38      this.preprocessingData = (ITransactionalPreprocessingData)preprocessingData;
39      originalColumns = new Dictionary<string, IList<double>>();
40      renamedColumns = new Dictionary<string, string>();
41    }
42
43    public bool ApplyTransformations(IEnumerable<ITransformation> transformations, bool preserveColumns, out string errorMsg) {
44      bool success;
45
46      preprocessingData.BeginTransaction(DataPreprocessingChangedEventType.Transformation);
47
48      try {
49        var doubleTransformations = transformations.OfType<Transformation<double>>().ToList();
50
51        if (preserveColumns) {
52          PreserveColumns(doubleTransformations);
53        }
54
55        // all transformations are performed inplace. no creation of new columns for transformations
56        ApplyDoubleTranformationsInplace(doubleTransformations, preserveColumns, out success, out errorMsg);
57
58        if (preserveColumns) {
59          RenameTransformedColumsAndRestorePreservedColumns(doubleTransformations);
60          RenameTransformationColumnParameter(doubleTransformations);
61          InsertCopyColumTransformations(doubleTransformations);
62
63          originalColumns.Clear();
64          renamedColumns.Clear();
65        }
66      }
67      finally {
68        preprocessingData.EndTransaction();
69      }
70
71      return success;
72    }
73
74    private void PreserveColumns(IEnumerable<Transformation<double>> transformations) {
75      foreach (var transformation in transformations) {
76        if (!originalColumns.ContainsKey(transformation.Column)) {
77          int colIndex = preprocessingData.GetColumnIndex(transformation.Column);
78          var originalData = preprocessingData.GetValues<double>(colIndex);
79          originalColumns.Add(transformation.Column, originalData);
80        }
81      }
82    }
83
84    private void ApplyDoubleTranformationsInplace(IEnumerable<Transformation<double>> transformations, bool preserveColumns, out bool success, out string errorMsg) {
85      errorMsg = string.Empty;
86      success = true;
87      foreach (var transformation in transformations) {
88        int colIndex = preprocessingData.GetColumnIndex(transformation.Column);
89
90        var originalData = preprocessingData.GetValues<double>(colIndex);
91
92        string errorMsgPart;
93        bool successPart;
94        var transformedData = ApplyDoubleTransformation(transformation, originalData, out successPart, out errorMsgPart);
95        errorMsg += errorMsgPart + Environment.NewLine;
96
97        if (!successPart) success = false;
98        preprocessingData.SetValues(colIndex, transformedData.ToList());
99        preprocessingData.Transformations.Add(transformation);
100      }
101    }
102
103    private IEnumerable<double> ApplyDoubleTransformation(Transformation<double> transformation, IList<double> data, out bool success, out string errorMsg) {
104      success = transformation.Check(data, out errorMsg);
105      return transformation.Apply(data);
106    }
107
108    private void RenameTransformationColumnParameter(List<Transformation<double>> transformations) {
109      foreach (var transformation in transformations) {
110        var newColumnName = new StringValue(renamedColumns[transformation.Column]);
111        transformation.ColumnParameter.ValidValues.Add(newColumnName);
112        transformation.ColumnParameter.Value = newColumnName;
113      }
114    }
115
116    private void InsertCopyColumTransformations(IList<Transformation<double>> transformations) {
117      foreach (var renaming in renamedColumns) {
118        string oldName = renaming.Key;
119        string newName = renaming.Value;
120
121        var copyTransformation = CreateCopyTransformation(oldName, newName);
122        preprocessingData.Transformations.Insert(0, copyTransformation);
123      }
124    }
125
126    private CopyColumnTransformation CreateCopyTransformation(string oldColumn, string newColumn) {
127      var newColumName = new StringValue(newColumn);
128
129      var copyTransformation = new CopyColumnTransformation();
130      copyTransformation.ColumnParameter.ValidValues.Add(newColumName);
131      copyTransformation.ColumnParameter.Value = newColumName;
132
133      copyTransformation.CopiedColumnNameParameter.Value.Value = oldColumn;
134      return copyTransformation;
135    }
136
137    private void RenameTransformedColumsAndRestorePreservedColumns(IList<Transformation<double>> transformations) {
138      foreach (var column in originalColumns) {
139        int originalColumnIndex = preprocessingData.GetColumnIndex(column.Key);
140        int newColumnIndex = originalColumnIndex + 1;
141        string newColumnName = GetTransformatedColumnName(transformations, column.Key);
142        // save renaming mapping
143        renamedColumns[column.Key] = newColumnName;
144        // create new transformed column
145        preprocessingData.InsertColumn<double>(newColumnName, newColumnIndex);
146        preprocessingData.SetValues(newColumnIndex, preprocessingData.GetValues<double>(originalColumnIndex));
147        // restore old values
148        preprocessingData.SetValues(originalColumnIndex, column.Value);
149      }
150    }
151
152    private string GetTransformatedColumnName(IList<Transformation<double>> transformations, string column) {
153      string suffix = GetTransformationSuffix(transformations, column);
154      return column + "_" + suffix;
155    }
156
157    private string GetTransformationSuffix(IList<Transformation<double>> transformations, string column) {
158      var suffixes = transformations.Where(t => t.Column == column).Select(t => t.ShortName);
159      var builder = new StringBuilder();
160      foreach (var suffix in suffixes) {
161        builder.Append(suffix);
162      }
163      return builder.ToString();
164    }
165  }
166}
Note: See TracBrowser for help on using the repository browser.