Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/sources/HeuristicLab.Problems.DataAnalysis/3.3/DataAnalysisProblemData.cs @ 4451

Last change on this file since 4451 was 4451, checked in by gkronber, 14 years ago

Made data analysis problem data and data analysis solutions savable. ##1193
File size: 17.5 KB
Line 
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
22using System;
23using System.Collections.Generic;
24using System.IO;
25using System.Linq;
26using HeuristicLab.Common;
27using HeuristicLab.Core;
28using HeuristicLab.Data;
29using HeuristicLab.Parameters;
30using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
31
32namespace HeuristicLab.Problems.DataAnalysis {
33  [Item("DataAnalysisProblemData", "Represents an item containing all data defining a data analysis problem.")]
34  [StorableClass]
35  public class DataAnalysisProblemData : ParameterizedNamedItem, IStorableContent {
36    protected 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    public IValueParameter<StringValue> TargetVariableParameter {
72      get { return (IValueParameter<StringValue>)Parameters["TargetVariable"]; }
73    }
74    public IValueParameter<ICheckedItemList<StringValue>> InputVariablesParameter {
75      get { return (IValueParameter<ICheckedItemList<StringValue>>)Parameters["InputVariables"]; }
76    }
77    public IValueParameter<IntValue> TrainingSamplesStartParameter {
78      get { return (IValueParameter<IntValue>)Parameters["TrainingSamplesStart"]; }
79    }
80    public IValueParameter<IntValue> TrainingSamplesEndParameter {
81      get { return (IValueParameter<IntValue>)Parameters["TrainingSamplesEnd"]; }
82    }
83    public IValueParameter<IntValue> TestSamplesStartParameter {
84      get { return (IValueParameter<IntValue>)Parameters["TestSamplesStart"]; }
85    }
86    public IValueParameter<IntValue> TestSamplesEndParameter {
87      get { return (IValueParameter<IntValue>)Parameters["TestSamplesEnd"]; }
88    }
89    #endregion
90
91    #region properties
92    public Dataset Dataset {
93      get { return (Dataset)DatasetParameter.Value; }
94      set {
95        if (value != Dataset) {
96          if (value == null) throw new ArgumentNullException();
97          DatasetParameter.Value = value;
98        }
99      }
100    }
101    public StringValue TargetVariable {
102      get { return (StringValue)TargetVariableParameter.Value; }
103      set {
104        if (value != TargetVariableParameter.Value) {
105          if (value == null) throw new ArgumentNullException();
106          if (TargetVariable != null) DeregisterStringValueEventHandlers(TargetVariable);
107          TargetVariableParameter.Value = value;
108        }
109      }
110    }
111    public ICheckedItemList<StringValue> InputVariables {
112      get { return (ICheckedItemList<StringValue>)InputVariablesParameter.Value; }
113      set {
114        if (value != InputVariables) {
115          if (value == null) throw new ArgumentNullException();
116          if (InputVariables != null) DeregisterInputVariablesEventHandlers();
117          InputVariablesParameter.Value = value;
118        }
119      }
120    }
121    public IntValue TrainingSamplesStart {
122      get { return (IntValue)TrainingSamplesStartParameter.Value; }
123      set {
124        if (value != TrainingSamplesStart) {
125          if (value == null) throw new ArgumentNullException();
126          if (TrainingSamplesStart != null) DeregisterValueTypeEventHandlers(TrainingSamplesStart);
127          TrainingSamplesStartParameter.Value = value;
128        }
129      }
130    }
131    public IntValue TrainingSamplesEnd {
132      get { return (IntValue)TrainingSamplesEndParameter.Value; }
133      set {
134        if (value != TrainingSamplesEnd) {
135          if (value == null) throw new ArgumentNullException();
136          if (TrainingSamplesEnd != null) DeregisterValueTypeEventHandlers(TrainingSamplesEnd);
137          TrainingSamplesEndParameter.Value = value;
138        }
139      }
140    }
141    public IntValue TestSamplesStart {
142      get { return (IntValue)TestSamplesStartParameter.Value; }
143      set {
144        if (value != TestSamplesStart) {
145          if (value == null) throw new ArgumentNullException();
146          if (TestSamplesStart != null) DeregisterValueTypeEventHandlers(TestSamplesStart);
147          TestSamplesStartParameter.Value = value;
148        }
149      }
150    }
151    public IntValue TestSamplesEnd {
152      get { return (IntValue)TestSamplesEndParameter.Value; }
153      set {
154        if (value != TestSamplesEnd) {
155          if (value == null) throw new ArgumentNullException();
156          if (TestSamplesEnd != null) DeregisterValueTypeEventHandlers(TestSamplesEnd);
157          TestSamplesEndParameter.Value = value;
158        }
159      }
160    }
161    #endregion
162
163    #region IStorableContent Members
164    public string Filename { get; set; }
165    #endregion
166
167    public DataAnalysisProblemData()
168      : base() {
169      var inputVariables = new CheckedItemList<StringValue>();
170      StringValue inputVariable = new StringValue("x");
171      inputVariables.Add(inputVariable);
172      StringValue targetVariable = new StringValue("y");
173      var validTargetVariables = new ItemSet<StringValue>();
174      validTargetVariables.Add(targetVariable);
175      Parameters.Add(new ValueParameter<Dataset>("Dataset", new Dataset(new string[] { "y", "x" }, kozaF1)));
176      Parameters.Add(new ValueParameter<ICheckedItemList<StringValue>>("InputVariables", inputVariables.AsReadOnly()));
177      Parameters.Add(new ConstrainedValueParameter<StringValue>("TargetVariable", validTargetVariables, targetVariable));
178      Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesStart", new IntValue(0)));
179      Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesEnd", new IntValue(15)));
180      Parameters.Add(new ValueParameter<IntValue>("TestSamplesStart", new IntValue(15)));
181      Parameters.Add(new ValueParameter<IntValue>("TestSamplesEnd", new IntValue(25)));
182      RegisterParameterEventHandlers();
183      RegisterParameterValueEventHandlers();
184    }
185
186    public DataAnalysisProblemData(Dataset dataset, IEnumerable<string> inputVariables, string targetVariable,
187      int trainingSamplesStart, int trainingSamplesEnd, int testSamplesStart, int testSamplesEnd) {
188      var inputVariablesList = new CheckedItemList<StringValue>(inputVariables.Select(x => new StringValue(x)).ToList());
189      StringValue targetVariableValue = new StringValue(targetVariable);
190      var validTargetVariables = new ItemSet<StringValue>();
191      foreach (var variable in dataset.VariableNames)
192        if (variable != targetVariable)
193          validTargetVariables.Add(new StringValue(variable));
194      validTargetVariables.Add(targetVariableValue);
195      Parameters.Add(new ValueParameter<Dataset>("Dataset", dataset));
196      Parameters.Add(new ValueParameter<ICheckedItemList<StringValue>>("InputVariables", inputVariablesList.AsReadOnly()));
197      Parameters.Add(new ConstrainedValueParameter<StringValue>("TargetVariable", validTargetVariables, targetVariableValue));
198      Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesStart", new IntValue(trainingSamplesStart)));
199      Parameters.Add(new ValueParameter<IntValue>("TrainingSamplesEnd", new IntValue(trainingSamplesEnd)));
200      Parameters.Add(new ValueParameter<IntValue>("TestSamplesStart", new IntValue(testSamplesStart)));
201      Parameters.Add(new ValueParameter<IntValue>("TestSamplesEnd", new IntValue(testSamplesEnd)));
202      RegisterParameterEventHandlers();
203      RegisterParameterValueEventHandlers();
204    }
205
206    [StorableConstructor]
207    protected DataAnalysisProblemData(bool deserializing) : base(deserializing) { }
208
209    [StorableHook(HookType.AfterDeserialization)]
210    private void AfterDeserializationHook() {
211      RegisterParameterEventHandlers();
212      RegisterParameterValueEventHandlers();
213    }
214
215    #region events
216    public event EventHandler ProblemDataChanged;
217    protected virtual void OnProblemDataChanged(EventArgs e) {
218      if (!suppressEvents) {
219        var listeners = ProblemDataChanged;
220        if (listeners != null) listeners(this, e);
221      }
222    }
223
224    private void RegisterParameterEventHandlers() {
225      DatasetParameter.ValueChanged += new EventHandler(DatasetParameter_ValueChanged);
226      InputVariablesParameter.ValueChanged += new EventHandler(InputVariablesParameter_ValueChanged);
227      TargetVariableParameter.ValueChanged += new EventHandler(TargetVariableParameter_ValueChanged);
228      TrainingSamplesStartParameter.ValueChanged += new EventHandler(TrainingSamplesStartParameter_ValueChanged);
229      TrainingSamplesEndParameter.ValueChanged += new EventHandler(TrainingSamplesEndParameter_ValueChanged);
230      TestSamplesStartParameter.ValueChanged += new EventHandler(TestSamplesStartParameter_ValueChanged);
231      TestSamplesEndParameter.ValueChanged += new EventHandler(TestSamplesEndParameter_ValueChanged);
232    }
233
234    private void RegisterParameterValueEventHandlers() {
235      RegisterInputVariablesEventHandlers();
236      if (TargetVariable != null) RegisterStringValueEventHandlers(TargetVariable);
237      RegisterValueTypeEventHandlers(TrainingSamplesStart);
238      RegisterValueTypeEventHandlers(TrainingSamplesEnd);
239      RegisterValueTypeEventHandlers(TestSamplesStart);
240      RegisterValueTypeEventHandlers(TestSamplesEnd);
241    }
242
243
244    #region parameter value changed event handlers
245    private void DatasetParameter_ValueChanged(object sender, EventArgs e) {
246      OnProblemDataChanged(EventArgs.Empty);
247    }
248    private void InputVariablesParameter_ValueChanged(object sender, EventArgs e) {
249      RegisterInputVariablesEventHandlers();
250      OnProblemDataChanged(EventArgs.Empty);
251    }
252    private void TargetVariableParameter_ValueChanged(object sender, EventArgs e) {
253      if (TargetVariable != null) {
254        RegisterStringValueEventHandlers(TargetVariable);
255        OnProblemDataChanged(EventArgs.Empty);
256      }
257    }
258    private void TrainingSamplesStartParameter_ValueChanged(object sender, EventArgs e) {
259      RegisterValueTypeEventHandlers(TrainingSamplesStart);
260      OnProblemDataChanged(EventArgs.Empty);
261    }
262    private void TrainingSamplesEndParameter_ValueChanged(object sender, EventArgs e) {
263      RegisterValueTypeEventHandlers(TrainingSamplesEnd);
264      OnProblemDataChanged(EventArgs.Empty);
265    }
266    private void TestSamplesStartParameter_ValueChanged(object sender, EventArgs e) {
267      RegisterValueTypeEventHandlers(TestSamplesStart);
268      OnProblemDataChanged(EventArgs.Empty);
269    }
270    private void TestSamplesEndParameter_ValueChanged(object sender, EventArgs e) {
271      RegisterValueTypeEventHandlers(TestSamplesEnd);
272      OnProblemDataChanged(EventArgs.Empty);
273    }
274    #endregion
275
276    private void RegisterInputVariablesEventHandlers() {
277      InputVariables.CollectionReset += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CollectionReset);
278      InputVariables.ItemsAdded += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsAdded);
279      InputVariables.ItemsRemoved += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsRemoved);
280      InputVariables.CheckedItemsChanged += new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CheckedItemsChanged);
281      foreach (var item in InputVariables) {
282        item.ValueChanged += new EventHandler(InputVariable_ValueChanged);
283      }
284    }
285
286    private void DeregisterInputVariablesEventHandlers() {
287      InputVariables.CollectionReset -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CollectionReset);
288      InputVariables.ItemsAdded -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsAdded);
289      InputVariables.ItemsRemoved -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_ItemsRemoved);
290      InputVariables.CheckedItemsChanged -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler<HeuristicLab.Collections.IndexedItem<StringValue>>(InputVariables_CheckedItemsChanged);
291      foreach (var item in InputVariables) {
292        item.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
293      }
294    }
295
296    private void InputVariables_CheckedItemsChanged(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs<HeuristicLab.Collections.IndexedItem<StringValue>> e) {
297      OnProblemDataChanged(e);
298    }
299    private void InputVariables_ItemsRemoved(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs<HeuristicLab.Collections.IndexedItem<StringValue>> e) {
300      foreach (var indexedItem in e.Items)
301        indexedItem.Value.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
302      OnProblemDataChanged(e);
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    private void InputVariables_CollectionReset(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs<HeuristicLab.Collections.IndexedItem<StringValue>> e) {
310      foreach (var indexedItem in e.OldItems)
311        indexedItem.Value.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
312      OnProblemDataChanged(e);
313    }
314    private void InputVariable_ValueChanged(object sender, EventArgs e) {
315      OnProblemDataChanged(e);
316    }
317
318    #region helper
319    private void RegisterValueTypeEventHandlers<T>(ValueTypeValue<T> value) where T : struct {
320      value.ValueChanged += new EventHandler(value_ValueChanged);
321    }
322    private void DeregisterValueTypeEventHandlers<T>(ValueTypeValue<T> value) where T : struct {
323      value.ValueChanged -= new EventHandler(value_ValueChanged);
324    }
325    private void RegisterStringValueEventHandlers(StringValue value) {
326      value.ValueChanged += new EventHandler(value_ValueChanged);
327    }
328    private void DeregisterStringValueEventHandlers(StringValue value) {
329      value.ValueChanged -= new EventHandler(value_ValueChanged);
330    }
331
332    private void value_ValueChanged(object sender, EventArgs e) {
333      OnProblemDataChanged(e);
334    }
335    #endregion
336    #endregion
337
338    public virtual void ImportFromFile(string fileName) {
339      var csvFileParser = new CsvFileParser();
340      csvFileParser.Parse(fileName);
341      suppressEvents = true;
342      Name = "Data imported from " + Path.GetFileName(fileName);
343      Dataset = new Dataset(csvFileParser.VariableNames, csvFileParser.Values);
344      Dataset.Name = Path.GetFileName(fileName);
345      var variableNames = Dataset.VariableNames.Select(x => new StringValue(x).AsReadOnly()).ToList();
346      ((ConstrainedValueParameter<StringValue>)TargetVariableParameter).ValidValues.Clear();
347      foreach (var variableName in variableNames)
348        ((ConstrainedValueParameter<StringValue>)TargetVariableParameter).ValidValues.Add(variableName);
349      TargetVariable = variableNames.First();
350      InputVariables = new CheckedItemList<StringValue>(variableNames).AsReadOnly();
351      InputVariables.SetItemCheckedState(variableNames.First(), false);
352      int middle = (int)(csvFileParser.Rows * 0.5);
353      TrainingSamplesStart = new IntValue(0);
354      TrainingSamplesEnd = new IntValue(middle);
355      TestSamplesStart = new IntValue(middle);
356      TestSamplesEnd = new IntValue(csvFileParser.Rows);
357      suppressEvents = false;
358      OnProblemDataChanged(EventArgs.Empty);
359    }
360
361    public override IDeepCloneable Clone(Cloner cloner) {
362      DataAnalysisProblemData clone = (DataAnalysisProblemData)base.Clone(cloner);
363      clone.RegisterParameterEventHandlers();
364      clone.RegisterParameterValueEventHandlers();
365      return clone;
366    }
367  }
368}
Note: See TracBrowser for help on using the repository browser.