1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 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.Collections.Generic;
|
---|
24 | using System.Linq;
|
---|
25 | using HeuristicLab.Common;
|
---|
26 | using HeuristicLab.Core;
|
---|
27 | using HeuristicLab.Data;
|
---|
28 | using HEAL.Attic;
|
---|
29 | using HeuristicLab.Problems.DataAnalysis;
|
---|
30 |
|
---|
31 | namespace HeuristicLab.DataPreprocessing {
|
---|
32 | [Item("FilteredPreprocessingData", "Represents filtered data used for preprocessing.")]
|
---|
33 | [StorableType("26BAE57C-A102-483D-8A09-AEC7132FD837")]
|
---|
34 | public sealed class FilteredPreprocessingData : NamedItem, IFilteredPreprocessingData {
|
---|
35 |
|
---|
36 | [Storable]
|
---|
37 | private readonly IPreprocessingData originalData;
|
---|
38 | [Storable]
|
---|
39 | private IPreprocessingData filteredData;
|
---|
40 |
|
---|
41 | public IPreprocessingData ActiveData {
|
---|
42 | get { return IsFiltered ? filteredData : originalData; }
|
---|
43 | }
|
---|
44 |
|
---|
45 | #region Constructor, Cloning & Persistence
|
---|
46 | public FilteredPreprocessingData(IPreprocessingData preprocessingData)
|
---|
47 | : base() {
|
---|
48 | originalData = preprocessingData;
|
---|
49 | filteredData = null;
|
---|
50 | }
|
---|
51 |
|
---|
52 | private FilteredPreprocessingData(FilteredPreprocessingData original, Cloner cloner)
|
---|
53 | : base(original, cloner) {
|
---|
54 | originalData = original.originalData;
|
---|
55 | filteredData = original.filteredData;
|
---|
56 | }
|
---|
57 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
58 | return new FilteredPreprocessingData(this, cloner);
|
---|
59 | }
|
---|
60 |
|
---|
61 | [StorableConstructor]
|
---|
62 | private FilteredPreprocessingData(StorableConstructorFlag _) : base(_) { }
|
---|
63 | #endregion
|
---|
64 |
|
---|
65 | #region Cells
|
---|
66 | public bool IsCellEmpty(int columnIndex, int rowIndex) {
|
---|
67 | return ActiveData.IsCellEmpty(columnIndex, rowIndex);
|
---|
68 | }
|
---|
69 |
|
---|
70 | public T GetCell<T>(int columnIndex, int rowIndex) {
|
---|
71 | return ActiveData.GetCell<T>(columnIndex, rowIndex);
|
---|
72 | }
|
---|
73 |
|
---|
74 | public void SetCell<T>(int columnIndex, int rowIndex, T value) {
|
---|
75 | if (IsFiltered)
|
---|
76 | throw new InvalidOperationException("SetValues not possible while data is filtered");
|
---|
77 | originalData.SetCell<T>(columnIndex, rowIndex, value);
|
---|
78 | }
|
---|
79 |
|
---|
80 | public string GetCellAsString(int columnIndex, int rowIndex) {
|
---|
81 | return ActiveData.GetCellAsString(columnIndex, rowIndex);
|
---|
82 | }
|
---|
83 |
|
---|
84 | public IList<T> GetValues<T>(int columnIndex, bool considerSelection) {
|
---|
85 | return ActiveData.GetValues<T>(columnIndex, considerSelection);
|
---|
86 | }
|
---|
87 |
|
---|
88 | public void SetValues<T>(int columnIndex, IList<T> values) {
|
---|
89 | if (IsFiltered)
|
---|
90 | throw new InvalidOperationException("SetValues not possible while data is filtered");
|
---|
91 |
|
---|
92 | originalData.SetValues<T>(columnIndex, values);
|
---|
93 | }
|
---|
94 |
|
---|
95 | public bool SetValue(string value, int columnIndex, int rowIndex) {
|
---|
96 | if (IsFiltered)
|
---|
97 | throw new InvalidOperationException("SetValue not possible while data is filtered");
|
---|
98 | return originalData.SetValue(value, columnIndex, rowIndex);
|
---|
99 | }
|
---|
100 |
|
---|
101 | public int Columns {
|
---|
102 | get { return ActiveData.Columns; }
|
---|
103 | }
|
---|
104 |
|
---|
105 | public int Rows {
|
---|
106 | get { return ActiveData.Rows; }
|
---|
107 | }
|
---|
108 | #endregion
|
---|
109 |
|
---|
110 | #region Rows
|
---|
111 | public void InsertRow(int rowIndex) {
|
---|
112 | if (IsFiltered)
|
---|
113 | throw new InvalidOperationException("InsertRow not possible while data is filtered");
|
---|
114 |
|
---|
115 | originalData.InsertRow(rowIndex);
|
---|
116 | }
|
---|
117 |
|
---|
118 | public void DeleteRow(int rowIndex) {
|
---|
119 | if (IsFiltered)
|
---|
120 | throw new InvalidOperationException("DeleteRow not possible while data is filtered");
|
---|
121 |
|
---|
122 | originalData.DeleteRow(rowIndex);
|
---|
123 | }
|
---|
124 |
|
---|
125 | public void DeleteRowsWithIndices(IEnumerable<int> rows) {
|
---|
126 | if (IsFiltered)
|
---|
127 | throw new InvalidOperationException("DeleteRowsWithIndices not possible while data is filtered");
|
---|
128 |
|
---|
129 | originalData.DeleteRowsWithIndices(rows);
|
---|
130 | }
|
---|
131 |
|
---|
132 | public void InsertColumn<T>(string variableName, int columnIndex) {
|
---|
133 | if (IsFiltered)
|
---|
134 | throw new InvalidOperationException("InsertColumn not possible while data is filtered");
|
---|
135 |
|
---|
136 | originalData.InsertColumn<T>(variableName, columnIndex);
|
---|
137 | }
|
---|
138 |
|
---|
139 | public void DeleteColumn(int columnIndex) {
|
---|
140 | if (IsFiltered)
|
---|
141 | throw new InvalidOperationException("DeleteColumn not possible while data is filtered");
|
---|
142 | originalData.DeleteColumn(columnIndex);
|
---|
143 | }
|
---|
144 |
|
---|
145 | public void RenameColumn(int columnIndex, string name) {
|
---|
146 | if (IsFiltered)
|
---|
147 | throw new InvalidOperationException("RenameColumn not possible while data is filtered");
|
---|
148 | originalData.RenameColumn(columnIndex, name);
|
---|
149 | }
|
---|
150 |
|
---|
151 | public void RenameColumns(IList<string> names) {
|
---|
152 | if (IsFiltered)
|
---|
153 | throw new InvalidOperationException("RenameColumns not possible while data is filtered");
|
---|
154 | originalData.RenameColumns(names);
|
---|
155 | }
|
---|
156 |
|
---|
157 | public bool AreAllStringColumns(IEnumerable<int> columnIndices) {
|
---|
158 | return originalData.AreAllStringColumns(columnIndices);
|
---|
159 | }
|
---|
160 | #endregion
|
---|
161 |
|
---|
162 | #region Variables
|
---|
163 | public IEnumerable<string> VariableNames {
|
---|
164 | get { return ActiveData.VariableNames; }
|
---|
165 | }
|
---|
166 | public IEnumerable<string> GetDoubleVariableNames() {
|
---|
167 | return originalData.GetDoubleVariableNames();
|
---|
168 | }
|
---|
169 | public string GetVariableName(int columnIndex) {
|
---|
170 | return ActiveData.GetVariableName(columnIndex);
|
---|
171 | }
|
---|
172 |
|
---|
173 | public int GetColumnIndex(string variableName) {
|
---|
174 | return ActiveData.GetColumnIndex(variableName);
|
---|
175 | }
|
---|
176 |
|
---|
177 | public bool VariableHasType<T>(int columnIndex) {
|
---|
178 | return originalData.VariableHasType<T>(columnIndex);
|
---|
179 | }
|
---|
180 |
|
---|
181 | public Type GetVariableType(int columnIndex) {
|
---|
182 | return ActiveData.GetVariableType(columnIndex);
|
---|
183 | }
|
---|
184 |
|
---|
185 | public IList<string> InputVariables {
|
---|
186 | get { return ActiveData.InputVariables; }
|
---|
187 | }
|
---|
188 |
|
---|
189 | public string TargetVariable {
|
---|
190 | get { return ActiveData.TargetVariable; }
|
---|
191 | } // optional
|
---|
192 | #endregion
|
---|
193 |
|
---|
194 | #region Partitions
|
---|
195 | public IntRange TrainingPartition {
|
---|
196 | get { return originalData.TrainingPartition; }
|
---|
197 | }
|
---|
198 |
|
---|
199 | public IntRange TestPartition {
|
---|
200 | get { return originalData.TestPartition; }
|
---|
201 | }
|
---|
202 | #endregion
|
---|
203 |
|
---|
204 | #region Transformations
|
---|
205 | public IList<ITransformation> Transformations {
|
---|
206 | get { return originalData.Transformations; }
|
---|
207 | }
|
---|
208 | #endregion
|
---|
209 |
|
---|
210 | #region Validation
|
---|
211 | public bool Validate(string value, out string errorMessage, int columnIndex) {
|
---|
212 | return originalData.Validate(value, out errorMessage, columnIndex);
|
---|
213 | }
|
---|
214 | #endregion
|
---|
215 |
|
---|
216 | #region Import & Export
|
---|
217 | public void Import(IDataAnalysisProblemData problemData) {
|
---|
218 | if (IsFiltered)
|
---|
219 | throw new InvalidOperationException("Import not possible while data is filtered");
|
---|
220 | originalData.Import(problemData);
|
---|
221 | }
|
---|
222 |
|
---|
223 | public Dataset ExportToDataset() {
|
---|
224 | return originalData.ExportToDataset();
|
---|
225 | }
|
---|
226 | #endregion
|
---|
227 |
|
---|
228 | #region Selection
|
---|
229 | public IDictionary<int, IList<int>> Selection {
|
---|
230 | get { return originalData.Selection; }
|
---|
231 | set { originalData.Selection = value; }
|
---|
232 | }
|
---|
233 |
|
---|
234 | public void ClearSelection() {
|
---|
235 | originalData.ClearSelection();
|
---|
236 | }
|
---|
237 |
|
---|
238 | public event EventHandler SelectionChanged {
|
---|
239 | add { originalData.SelectionChanged += value; }
|
---|
240 | remove { originalData.SelectionChanged -= value; }
|
---|
241 | }
|
---|
242 | #endregion
|
---|
243 |
|
---|
244 | #region Transactions
|
---|
245 | public event DataPreprocessingChangedEventHandler Changed {
|
---|
246 | add { originalData.Changed += value; }
|
---|
247 | remove { originalData.Changed -= value; }
|
---|
248 | }
|
---|
249 |
|
---|
250 | public bool IsUndoAvailable {
|
---|
251 | get { return IsFiltered ? false : originalData.IsUndoAvailable; }
|
---|
252 | }
|
---|
253 |
|
---|
254 | public void Undo() {
|
---|
255 | if (IsFiltered)
|
---|
256 | throw new InvalidOperationException("Undo not possible while data is filtered");
|
---|
257 |
|
---|
258 | originalData.Undo();
|
---|
259 | }
|
---|
260 |
|
---|
261 | public void InTransaction(Action action, DataPreprocessingChangedEventType type = DataPreprocessingChangedEventType.Any) {
|
---|
262 | if (IsFiltered)
|
---|
263 | throw new InvalidOperationException("Transaction not possible while data is filtered");
|
---|
264 | originalData.InTransaction(action, type);
|
---|
265 | }
|
---|
266 |
|
---|
267 | public void BeginTransaction(DataPreprocessingChangedEventType type) {
|
---|
268 | if (IsFiltered)
|
---|
269 | throw new InvalidOperationException("Transaction not possible while data is filtered");
|
---|
270 | originalData.BeginTransaction(type);
|
---|
271 | }
|
---|
272 |
|
---|
273 | public void EndTransaction() {
|
---|
274 | originalData.EndTransaction();
|
---|
275 | }
|
---|
276 | #endregion
|
---|
277 |
|
---|
278 | #region Statistics
|
---|
279 | public T GetMin<T>(int columnIndex, bool considerSelection = false, T emptyValue = default(T)) {
|
---|
280 | return ActiveData.GetMin<T>(columnIndex, considerSelection, emptyValue);
|
---|
281 | }
|
---|
282 | public T GetMax<T>(int columnIndex, bool considerSelection = false, T emptyValue = default(T)) {
|
---|
283 | return ActiveData.GetMax<T>(columnIndex, considerSelection, emptyValue);
|
---|
284 | }
|
---|
285 | public T GetMean<T>(int columnIndex, bool considerSelection = false, T emptyValue = default(T)) {
|
---|
286 | return ActiveData.GetMean<T>(columnIndex, considerSelection, emptyValue);
|
---|
287 | }
|
---|
288 | public T GetMedian<T>(int columnIndex, bool considerSelection = false, T emptyValue = default(T)) where T : IComparable<T> {
|
---|
289 | return ActiveData.GetMedian<T>(columnIndex, considerSelection, emptyValue);
|
---|
290 | }
|
---|
291 | public T GetMode<T>(int columnIndex, bool considerSelection = false, T emptyValue = default(T)) where T : IEquatable<T> {
|
---|
292 | return ActiveData.GetMode<T>(columnIndex, considerSelection, emptyValue);
|
---|
293 | }
|
---|
294 | public T GetStandardDeviation<T>(int columnIndex, bool considerSelection = false, T emptyValue = default(T)) {
|
---|
295 | return ActiveData.GetStandardDeviation<T>(columnIndex, considerSelection, emptyValue);
|
---|
296 | }
|
---|
297 | public T GetVariance<T>(int columnIndex, bool considerSelection = false, T emptyValue = default(T)) {
|
---|
298 | return ActiveData.GetVariance<T>(columnIndex, considerSelection, emptyValue);
|
---|
299 | }
|
---|
300 | public T GetQuantile<T>(double alpha, int columnIndex, bool considerSelection = false, T emptyValue = default(T)) where T : IComparable<T> {
|
---|
301 | return ActiveData.GetQuantile<T>(alpha, columnIndex, considerSelection, emptyValue);
|
---|
302 | }
|
---|
303 | public int GetDistinctValues<T>(int columnIndex, bool considerSelection = false) {
|
---|
304 | return ActiveData.GetDistinctValues<T>(columnIndex, considerSelection);
|
---|
305 | }
|
---|
306 |
|
---|
307 | public int GetMissingValueCount() {
|
---|
308 | return ActiveData.GetMissingValueCount();
|
---|
309 | }
|
---|
310 | public int GetMissingValueCount(int columnIndex) {
|
---|
311 | return ActiveData.GetMissingValueCount(columnIndex);
|
---|
312 | }
|
---|
313 | public int GetRowMissingValueCount(int rowIndex) {
|
---|
314 | return ActiveData.GetRowMissingValueCount(rowIndex);
|
---|
315 | }
|
---|
316 | #endregion
|
---|
317 |
|
---|
318 | #region Filters
|
---|
319 | public void SetFilter(bool[] remainingRows) {
|
---|
320 | filteredData = (IPreprocessingData)originalData.Clone();
|
---|
321 | filteredData.InTransaction(() => {
|
---|
322 | var remainingIndices = Enumerable.Range(0, remainingRows.Length).Where(x => remainingRows[x]);
|
---|
323 |
|
---|
324 | foreach (var v in filteredData.VariableNames) {
|
---|
325 | var ci = filteredData.GetColumnIndex(v);
|
---|
326 | if (filteredData.VariableHasType<double>(ci)) {
|
---|
327 | var values = filteredData.GetValues<double>(ci);
|
---|
328 | var filteredValues = remainingIndices.Select(x => values[x]).ToList();
|
---|
329 | filteredData.SetValues(ci, filteredValues);
|
---|
330 | } else if (filteredData.VariableHasType<DateTime>(ci)) {
|
---|
331 | var values = filteredData.GetValues<DateTime>(ci);
|
---|
332 | var filteredValues = remainingIndices.Select(x => values[x]).ToList();
|
---|
333 | filteredData.SetValues(ci, filteredValues);
|
---|
334 | } else if (filteredData.VariableHasType<string>(ci)) {
|
---|
335 | var values = filteredData.GetValues<string>(ci);
|
---|
336 | var filteredValues = remainingIndices.Select(x => values[x]).ToList();
|
---|
337 | filteredData.SetValues(ci, filteredValues);
|
---|
338 | }
|
---|
339 | }
|
---|
340 | });
|
---|
341 | OnFilterChanged();
|
---|
342 | }
|
---|
343 |
|
---|
344 | public void PersistFilter() {
|
---|
345 | originalData.InTransaction(() => {
|
---|
346 | for (int i = 0; i < filteredData.Columns; ++i) {
|
---|
347 | if (filteredData.VariableHasType<double>(i)) {
|
---|
348 | originalData.SetValues<double>(i, filteredData.GetValues<double>(i));
|
---|
349 | } else if (filteredData.VariableHasType<string>(i)) {
|
---|
350 | originalData.SetValues<string>(i, filteredData.GetValues<string>(i));
|
---|
351 | } else if (filteredData.VariableHasType<DateTime>(i)) {
|
---|
352 | originalData.SetValues<DateTime>(i, filteredData.GetValues<DateTime>(i));
|
---|
353 | } else {
|
---|
354 | throw new ArgumentException("Data types of columns do not match");
|
---|
355 | }
|
---|
356 | }
|
---|
357 | });
|
---|
358 | ResetFilter();
|
---|
359 | }
|
---|
360 |
|
---|
361 | public void ResetFilter() {
|
---|
362 | filteredData = null;
|
---|
363 | OnFilterChanged();
|
---|
364 | }
|
---|
365 |
|
---|
366 | public bool IsFiltered {
|
---|
367 | get { return filteredData != null; }
|
---|
368 | }
|
---|
369 |
|
---|
370 | public event EventHandler FilterChanged;
|
---|
371 |
|
---|
372 | private void OnFilterChanged() {
|
---|
373 | if (FilterChanged != null) {
|
---|
374 | FilterChanged(this, new EventArgs());
|
---|
375 | }
|
---|
376 | }
|
---|
377 | #endregion
|
---|
378 | }
|
---|
379 | }
|
---|