1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2010 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.IO;
|
---|
25 | using System.Linq;
|
---|
26 | using HeuristicLab.Common;
|
---|
27 | using HeuristicLab.Core;
|
---|
28 | using HeuristicLab.Data;
|
---|
29 | using HeuristicLab.Parameters;
|
---|
30 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
31 |
|
---|
32 | namespace HeuristicLab.Problems.DataAnalysis {
|
---|
33 | [Item("DataAnalysisProblemData", "Represents an item containing all data defining a data analysis problem.")]
|
---|
34 | [StorableClass]
|
---|
35 | public class DataAnalysisProblemData : ParameterizedNamedItem {
|
---|
36 | private bool suppressEvents = false;
|
---|
37 | #region default data
|
---|
38 | // y = x^4 + x^3 + x^2 + x
|
---|
39 | private static double[,] kozaF1 = new double[,] {
|
---|
40 | {2.017885919, -1.449165046},
|
---|
41 | {1.30060506, -1.344523885},
|
---|
42 | {1.147134798, -1.317989331},
|
---|
43 | {0.877182504, -1.266142284},
|
---|
44 | {0.852562452, -1.261020794},
|
---|
45 | {0.431095788, -1.158793317},
|
---|
46 | {0.112586002, -1.050908405},
|
---|
47 | {0.04594507, -1.021989402},
|
---|
48 | {0.042572879, -1.020438113},
|
---|
49 | {-0.074027291, -0.959859562},
|
---|
50 | {-0.109178553, -0.938094706},
|
---|
51 | {-0.259721109, -0.803635355},
|
---|
52 | {-0.272991057, -0.387519561},
|
---|
53 | {-0.161978191, -0.193611001},
|
---|
54 | {-0.102489983, -0.114215349},
|
---|
55 | {-0.01469968, -0.014918985},
|
---|
56 | {-0.008863365, -0.008942626},
|
---|
57 | {0.026751057, 0.026054094},
|
---|
58 | {0.166922436, 0.14309643},
|
---|
59 | {0.176953808, 0.1504144},
|
---|
60 | {0.190233418, 0.159916534},
|
---|
61 | {0.199800708, 0.166635331},
|
---|
62 | {0.261502822, 0.207600348},
|
---|
63 | {0.30182879, 0.232370249},
|
---|
64 | {0.83763905, 0.468046718}
|
---|
65 | };
|
---|
66 | #endregion
|
---|
67 | #region parameter properties
|
---|
68 | public IValueParameter<Dataset> DatasetParameter {
|
---|
69 | get { return (IValueParameter<Dataset>)Parameters["Dataset"]; }
|
---|
70 | }
|
---|
71 |
|
---|
72 | public IValueParameter<StringValue> TargetVariableParameter {
|
---|
73 | get { return (IValueParameter<StringValue>)Parameters["TargetVariable"]; }
|
---|
74 | }
|
---|
75 |
|
---|
76 | public IValueParameter<ICheckedItemList<StringValue>> InputVariablesParameter {
|
---|
77 | get { return (IValueParameter<ICheckedItemList<StringValue>>)Parameters["InputVariables"]; }
|
---|
78 | }
|
---|
79 |
|
---|
80 | public IValueParameter<IntValue> TrainingSamplesStartParameter {
|
---|
81 | get { return (IValueParameter<IntValue>)Parameters["TrainingSamplesStart"]; }
|
---|
82 | }
|
---|
83 | public IValueParameter<IntValue> TrainingSamplesEndParameter {
|
---|
84 | get { return (IValueParameter<IntValue>)Parameters["TrainingSamplesEnd"]; }
|
---|
85 | }
|
---|
86 | public IValueParameter<IntValue> TestSamplesStartParameter {
|
---|
87 | get { return (IValueParameter<IntValue>)Parameters["TestSamplesStart"]; }
|
---|
88 | }
|
---|
89 | public IValueParameter<IntValue> TestSamplesEndParameter {
|
---|
90 | get { return (IValueParameter<IntValue>)Parameters["TestSamplesEnd"]; }
|
---|
91 | }
|
---|
92 | #endregion
|
---|
93 |
|
---|
94 | #region properties
|
---|
95 | public Dataset Dataset {
|
---|
96 | get { return (Dataset)DatasetParameter.Value; }
|
---|
97 | set {
|
---|
98 | if (value != Dataset) {
|
---|
99 | if (value == null) throw new ArgumentNullException();
|
---|
100 | DatasetParameter.Value = value;
|
---|
101 | }
|
---|
102 | }
|
---|
103 | }
|
---|
104 | public StringValue TargetVariable {
|
---|
105 | get { return (StringValue)TargetVariableParameter.Value; }
|
---|
106 | set {
|
---|
107 | if (value != TargetVariableParameter.Value) {
|
---|
108 | if (value == null) throw new ArgumentNullException();
|
---|
109 | if (TargetVariable != null) DeregisterStringValueEventHandlers(TargetVariable);
|
---|
110 | TargetVariableParameter.Value = value;
|
---|
111 | }
|
---|
112 | }
|
---|
113 | }
|
---|
114 | public ICheckedItemList<StringValue> InputVariables {
|
---|
115 | get { return (ICheckedItemList<StringValue>)InputVariablesParameter.Value; }
|
---|
116 | set {
|
---|
117 | if (value != InputVariables) {
|
---|
118 | if (value == null) throw new ArgumentNullException();
|
---|
119 | if (InputVariables != null) DeregisterInputVariablesEventHandlers();
|
---|
120 | InputVariablesParameter.Value = value;
|
---|
121 | }
|
---|
122 | }
|
---|
123 | }
|
---|
124 | public IntValue TrainingSamplesStart {
|
---|
125 | get { return (IntValue)TrainingSamplesStartParameter.Value; }
|
---|
126 | set {
|
---|
127 | if (value != TrainingSamplesStart) {
|
---|
128 | if (value == null) throw new ArgumentNullException();
|
---|
129 | if (TrainingSamplesStart != null) DeregisterValueTypeEventHandlers(TrainingSamplesStart);
|
---|
130 | TrainingSamplesStartParameter.Value = value;
|
---|
131 | }
|
---|
132 | }
|
---|
133 | }
|
---|
134 | public IntValue TrainingSamplesEnd {
|
---|
135 | get { return (IntValue)TrainingSamplesEndParameter.Value; }
|
---|
136 | set {
|
---|
137 | if (value != TrainingSamplesEnd) {
|
---|
138 | if (value == null) throw new ArgumentNullException();
|
---|
139 | if (TrainingSamplesEnd != null) DeregisterValueTypeEventHandlers(TrainingSamplesEnd);
|
---|
140 | TrainingSamplesEndParameter.Value = value;
|
---|
141 | }
|
---|
142 | }
|
---|
143 | }
|
---|
144 | public IntValue TestSamplesStart {
|
---|
145 | get { return (IntValue)TestSamplesStartParameter.Value; }
|
---|
146 | set {
|
---|
147 | if (value != TestSamplesStart) {
|
---|
148 | if (value == null) throw new ArgumentNullException();
|
---|
149 | if (TestSamplesStart != null) DeregisterValueTypeEventHandlers(TestSamplesStart);
|
---|
150 | TestSamplesStartParameter.Value = value;
|
---|
151 | }
|
---|
152 | }
|
---|
153 | }
|
---|
154 | public IntValue TestSamplesEnd {
|
---|
155 | get { return (IntValue)TestSamplesEndParameter.Value; }
|
---|
156 | set {
|
---|
157 | if (value != TestSamplesEnd) {
|
---|
158 | if (value == null) throw new ArgumentNullException();
|
---|
159 | if (TestSamplesEnd != null) DeregisterValueTypeEventHandlers(TestSamplesEnd);
|
---|
160 | TestSamplesEndParameter.Value = value;
|
---|
161 | }
|
---|
162 | }
|
---|
163 | }
|
---|
164 | #endregion
|
---|
165 |
|
---|
166 | public DataAnalysisProblemData()
|
---|
167 | : base() {
|
---|
168 | var inputVariables = new CheckedItemList<StringValue>();
|
---|
169 | StringValue inputVariable = new StringValue("x");
|
---|
170 | inputVariables.Add(inputVariable);
|
---|
171 | StringValue targetVariable = new StringValue("y");
|
---|
172 | var validTargetVariables = new ItemSet<StringValue>();
|
---|
173 | validTargetVariables.Add(targetVariable);
|
---|
174 | Parameters.Add(new ValueParameter<Dataset>("Dataset", new Dataset(new string[] { "y", "x" }, kozaF1)));
|
---|
175 | Parameters.Add(new ValueParameter<ICheckedItemList<StringValue>>("InputVariables", inputVariables.AsReadOnly()));
|
---|
176 | Parameters.Add(new ConstrainedValueParameter<StringValue>("TargetVariable", validTargetVariables, targetVariable));
|
---|
177 | Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesStart", new IntValue(0)));
|
---|
178 | Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesEnd", new IntValue(15)));
|
---|
179 | Parameters.Add(new ValueParameter<IntValue>("TestSamplesStart", new IntValue(15)));
|
---|
180 | Parameters.Add(new ValueParameter<IntValue>("TestSamplesEnd", new IntValue(25)));
|
---|
181 | RegisterParameterEventHandlers();
|
---|
182 | RegisterParameterValueEventHandlers();
|
---|
183 | }
|
---|
184 |
|
---|
185 | public DataAnalysisProblemData(Dataset dataset, IEnumerable<string> inputVariables, string targetVariable,
|
---|
186 | int trainingSamplesStart, int trainingSamplesEnd, int testSamplesStart, int testSamplesEnd) {
|
---|
187 | var inputVariablesList = new CheckedItemList<StringValue>(inputVariables.Select(x => new StringValue(x)));
|
---|
188 | StringValue targetVariableValue = new StringValue(targetVariable);
|
---|
189 | var validTargetVariables = new ItemSet<StringValue>();
|
---|
190 | foreach (var variable in dataset.VariableNames)
|
---|
191 | if (variable != targetVariable)
|
---|
192 | validTargetVariables.Add(new StringValue(variable));
|
---|
193 | validTargetVariables.Add(targetVariableValue);
|
---|
194 | Parameters.Add(new ValueParameter<Dataset>("Dataset", dataset));
|
---|
195 | Parameters.Add(new ValueParameter<ICheckedItemList<StringValue>>("InputVariables", inputVariablesList.AsReadOnly()));
|
---|
196 | Parameters.Add(new ConstrainedValueParameter<StringValue>("TargetVariable", validTargetVariables, targetVariableValue));
|
---|
197 | Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesStart", new IntValue(trainingSamplesStart)));
|
---|
198 | Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesEnd", new IntValue(trainingSamplesEnd)));
|
---|
199 | Parameters.Add(new ValueParameter<IntValue>("TestSamplesStart", new IntValue(testSamplesStart)));
|
---|
200 | Parameters.Add(new ValueParameter<IntValue>("TestSamplesEnd", new IntValue(testSamplesEnd)));
|
---|
201 | RegisterParameterEventHandlers();
|
---|
202 | RegisterParameterValueEventHandlers();
|
---|
203 | }
|
---|
204 |
|
---|
205 | [StorableConstructor]
|
---|
206 | private DataAnalysisProblemData(bool deserializing) : base() { }
|
---|
207 |
|
---|
208 | [StorableHook(HookType.AfterDeserialization)]
|
---|
209 | private void AfterDeserializationHook() {
|
---|
210 | RegisterParameterEventHandlers();
|
---|
211 | RegisterParameterValueEventHandlers();
|
---|
212 | }
|
---|
213 |
|
---|
214 | #region events
|
---|
215 | public event EventHandler ProblemDataChanged;
|
---|
216 | protected virtual void OnProblemDataChanged(EventArgs e) {
|
---|
217 | if (!suppressEvents) {
|
---|
218 | var listeners = ProblemDataChanged;
|
---|
219 | if (listeners != null) listeners(this, e);
|
---|
220 | }
|
---|
221 | }
|
---|
222 |
|
---|
223 | private void RegisterParameterEventHandlers() {
|
---|
224 | DatasetParameter.ValueChanged += new EventHandler(DatasetParameter_ValueChanged);
|
---|
225 | InputVariablesParameter.ValueChanged += new EventHandler(InputVariablesParameter_ValueChanged);
|
---|
226 | TargetVariableParameter.ValueChanged += new EventHandler(TargetVariableParameter_ValueChanged);
|
---|
227 | TrainingSamplesStartParameter.ValueChanged += new EventHandler(TrainingSamplesStartParameter_ValueChanged);
|
---|
228 | TrainingSamplesEndParameter.ValueChanged += new EventHandler(TrainingSamplesEndParameter_ValueChanged);
|
---|
229 | TestSamplesStartParameter.ValueChanged += new EventHandler(TestSamplesStartParameter_ValueChanged);
|
---|
230 | TestSamplesEndParameter.ValueChanged += new EventHandler(TestSamplesEndParameter_ValueChanged);
|
---|
231 | }
|
---|
232 |
|
---|
233 | private void RegisterParameterValueEventHandlers() {
|
---|
234 | RegisterInputVariablesEventHandlers();
|
---|
235 | if (TargetVariable != null) RegisterStringValueEventHandlers(TargetVariable);
|
---|
236 | RegisterValueTypeEventHandlers(TrainingSamplesStart);
|
---|
237 | RegisterValueTypeEventHandlers(TrainingSamplesEnd);
|
---|
238 | RegisterValueTypeEventHandlers(TestSamplesStart);
|
---|
239 | RegisterValueTypeEventHandlers(TestSamplesEnd);
|
---|
240 | }
|
---|
241 |
|
---|
242 |
|
---|
243 | #region parameter value changed event handlers
|
---|
244 | void DatasetParameter_ValueChanged(object sender, EventArgs e) {
|
---|
245 | OnProblemDataChanged(EventArgs.Empty);
|
---|
246 | }
|
---|
247 | void InputVariablesParameter_ValueChanged(object sender, EventArgs e) {
|
---|
248 | RegisterInputVariablesEventHandlers();
|
---|
249 | OnProblemDataChanged(EventArgs.Empty);
|
---|
250 | }
|
---|
251 | void TargetVariableParameter_ValueChanged(object sender, EventArgs e) {
|
---|
252 | if (TargetVariable != null) {
|
---|
253 | RegisterStringValueEventHandlers(TargetVariable);
|
---|
254 | OnProblemDataChanged(EventArgs.Empty);
|
---|
255 | }
|
---|
256 | }
|
---|
257 | void TrainingSamplesStartParameter_ValueChanged(object sender, EventArgs e) {
|
---|
258 | RegisterValueTypeEventHandlers(TrainingSamplesStart);
|
---|
259 | OnProblemDataChanged(EventArgs.Empty);
|
---|
260 | }
|
---|
261 | void TrainingSamplesEndParameter_ValueChanged(object sender, EventArgs e) {
|
---|
262 | RegisterValueTypeEventHandlers(TrainingSamplesEnd);
|
---|
263 | OnProblemDataChanged(EventArgs.Empty);
|
---|
264 | }
|
---|
265 | void TestSamplesStartParameter_ValueChanged(object sender, EventArgs e) {
|
---|
266 | RegisterValueTypeEventHandlers(TestSamplesStart);
|
---|
267 | OnProblemDataChanged(EventArgs.Empty);
|
---|
268 | }
|
---|
269 | void TestSamplesEndParameter_ValueChanged(object sender, EventArgs e) {
|
---|
270 | RegisterValueTypeEventHandlers(TestSamplesEnd);
|
---|
271 | OnProblemDataChanged(EventArgs.Empty);
|
---|
272 | }
|
---|
273 | #endregion
|
---|
274 |
|
---|
275 | private void RegisterInputVariablesEventHandlers() {
|
---|
276 | InputVariables.CollectionReset += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CollectionReset);
|
---|
277 | InputVariables.ItemsAdded += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsAdded);
|
---|
278 | InputVariables.ItemsRemoved += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsRemoved);
|
---|
279 | InputVariables.CheckedItemsChanged += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CheckedItemsChanged);
|
---|
280 | foreach (var item in InputVariables)
|
---|
281 | item.ValueChanged += new EventHandler(InputVariable_ValueChanged);
|
---|
282 | }
|
---|
283 |
|
---|
284 | private void DeregisterInputVariablesEventHandlers() {
|
---|
285 | InputVariables.CollectionReset -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CollectionReset);
|
---|
286 | InputVariables.ItemsAdded -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsAdded);
|
---|
287 | InputVariables.ItemsRemoved -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsRemoved);
|
---|
288 | InputVariables.CheckedItemsChanged -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CheckedItemsChanged);
|
---|
289 | foreach (var item in InputVariables) {
|
---|
290 | item.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
|
---|
291 | }
|
---|
292 | }
|
---|
293 |
|
---|
294 | private void InputVariables_CheckedItemsChanged(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs<HeuristicLab.Collections.IndexedItem<StringValue>> e) {
|
---|
295 | OnProblemDataChanged(e);
|
---|
296 | }
|
---|
297 |
|
---|
298 | private void InputVariables_ItemsRemoved(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs<HeuristicLab.Collections.IndexedItem<StringValue>> e) {
|
---|
299 | foreach (var indexedItem in e.Items)
|
---|
300 | indexedItem.Value.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
|
---|
301 | OnProblemDataChanged(e);
|
---|
302 | }
|
---|
303 |
|
---|
304 | private void InputVariables_ItemsAdded(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs<HeuristicLab.Collections.IndexedItem<StringValue>> e) {
|
---|
305 | foreach (var indexedItem in e.Items)
|
---|
306 | indexedItem.Value.ValueChanged += new EventHandler(InputVariable_ValueChanged);
|
---|
307 | OnProblemDataChanged(e);
|
---|
308 | }
|
---|
309 |
|
---|
310 | private void InputVariables_CollectionReset(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs<HeuristicLab.Collections.IndexedItem<StringValue>> e) {
|
---|
311 | foreach (var indexedItem in e.OldItems)
|
---|
312 | indexedItem.Value.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
|
---|
313 | OnProblemDataChanged(e);
|
---|
314 | }
|
---|
315 |
|
---|
316 | void InputVariable_ValueChanged(object sender, EventArgs e) {
|
---|
317 | OnProblemDataChanged(e);
|
---|
318 | }
|
---|
319 | #region helper
|
---|
320 |
|
---|
321 | private void RegisterValueTypeEventHandlers<T>(ValueTypeValue<T> value) where T : struct {
|
---|
322 | value.ValueChanged += new EventHandler(value_ValueChanged);
|
---|
323 | }
|
---|
324 |
|
---|
325 | private void DeregisterValueTypeEventHandlers<T>(ValueTypeValue<T> value) where T : struct {
|
---|
326 | value.ValueChanged -= new EventHandler(value_ValueChanged);
|
---|
327 | }
|
---|
328 |
|
---|
329 | void value_ValueChanged(object sender, EventArgs e) {
|
---|
330 | OnProblemDataChanged(e);
|
---|
331 | }
|
---|
332 |
|
---|
333 | private void RegisterStringValueEventHandlers(StringValue value) {
|
---|
334 | value.ValueChanged += new EventHandler(value_ValueChanged);
|
---|
335 | }
|
---|
336 |
|
---|
337 | private void DeregisterStringValueEventHandlers(StringValue value) {
|
---|
338 | value.ValueChanged -= new EventHandler(value_ValueChanged);
|
---|
339 | }
|
---|
340 |
|
---|
341 | #endregion
|
---|
342 | #endregion
|
---|
343 |
|
---|
344 | public virtual void ImportFromFile(string fileName) {
|
---|
345 | var csvFileParser = new CsvFileParser();
|
---|
346 | csvFileParser.Parse(fileName);
|
---|
347 | suppressEvents = true;
|
---|
348 | Name = "Data imported from " + Path.GetFileName(fileName);
|
---|
349 | Dataset = new Dataset(csvFileParser.VariableNames, csvFileParser.Values);
|
---|
350 | Dataset.Name = Path.GetFileName(fileName);
|
---|
351 | var variableNames = Dataset.VariableNames.Select(x => new StringValue(x).AsReadOnly()).ToList();
|
---|
352 | ((ConstrainedValueParameter<StringValue>)TargetVariableParameter).ValidValues.Clear();
|
---|
353 | foreach (var variableName in variableNames)
|
---|
354 | ((ConstrainedValueParameter<StringValue>)TargetVariableParameter).ValidValues.Add(variableName);
|
---|
355 | TargetVariable = variableNames.First();
|
---|
356 | InputVariables = new CheckedItemList<StringValue>(variableNames).AsReadOnly();
|
---|
357 | InputVariables.SetItemCheckedState(variableNames.First(), false);
|
---|
358 | int middle = (int)(csvFileParser.Rows * 0.5);
|
---|
359 | TrainingSamplesStart = new IntValue(0);
|
---|
360 | TrainingSamplesEnd = new IntValue(middle);
|
---|
361 | TestSamplesStart = new IntValue(middle);
|
---|
362 | TestSamplesEnd = new IntValue(csvFileParser.Rows);
|
---|
363 | suppressEvents = false;
|
---|
364 | OnProblemDataChanged(EventArgs.Empty);
|
---|
365 | }
|
---|
366 |
|
---|
367 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
368 | DataAnalysisProblemData clone = (DataAnalysisProblemData)base.Clone(cloner);
|
---|
369 | clone.RegisterParameterEventHandlers();
|
---|
370 | clone.RegisterParameterValueEventHandlers();
|
---|
371 | return clone;
|
---|
372 | }
|
---|
373 | }
|
---|
374 | }
|
---|