Free cookie consent management tool by TermsFeed Policy Generator

source: branches/2839_HiveProjectManagement/HeuristicLab.Analysis/3.3/AlleleFrequencyAnalysis/AlleleFrequencyAnalyzer.cs @ 18066

Last change on this file since 18066 was 16057, checked in by jkarder, 6 years ago

#2839:
File size: 16.5 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2018 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 HeuristicLab.Common;
26using HeuristicLab.Core;
27using HeuristicLab.Data;
28using HeuristicLab.Operators;
29using HeuristicLab.Optimization;
30using HeuristicLab.Parameters;
31using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
32
33namespace HeuristicLab.Analysis {
34  /// <summary>
35  /// An operator for analyzing the frequency of alleles.
36  /// </summary>
37  [Item("AlleleFrequencyAnalyzer", "An operator for analyzing the frequency of alleles.")]
38  [StorableClass]
39  public abstract class AlleleFrequencyAnalyzer<T> : SingleSuccessorOperator, IAnalyzer, ISingleObjectiveOperator where T : class, IItem {
40    public virtual bool EnabledByDefault {
41      get { return false; }
42    }
43
44    public LookupParameter<BoolValue> MaximizationParameter {
45      get { return (LookupParameter<BoolValue>)Parameters["Maximization"]; }
46    }
47    public ScopeTreeLookupParameter<T> SolutionParameter {
48      get { return (ScopeTreeLookupParameter<T>)Parameters["Solution"]; }
49    }
50    public ScopeTreeLookupParameter<DoubleValue> QualityParameter {
51      get { return (ScopeTreeLookupParameter<DoubleValue>)Parameters["Quality"]; }
52    }
53    public LookupParameter<T> BestKnownSolutionParameter {
54      get { return (LookupParameter<T>)Parameters["BestKnownSolution"]; }
55    }
56    public ValueLookupParameter<ResultCollection> ResultsParameter {
57      get { return (ValueLookupParameter<ResultCollection>)Parameters["Results"]; }
58    }
59    public ValueParameter<BoolValue> StoreHistoryParameter {
60      get { return (ValueParameter<BoolValue>)Parameters["StoreHistory"]; }
61    }
62    public ValueParameter<IntValue> UpdateIntervalParameter {
63      get { return (ValueParameter<IntValue>)Parameters["UpdateInterval"]; }
64    }
65    public LookupParameter<IntValue> UpdateCounterParameter {
66      get { return (LookupParameter<IntValue>)Parameters["UpdateCounter"]; }
67    }
68
69    [StorableConstructor]
70    protected AlleleFrequencyAnalyzer(bool deserializing) : base(deserializing) { }
71    protected AlleleFrequencyAnalyzer(AlleleFrequencyAnalyzer<T> original, Cloner cloner) : base(original, cloner) { }
72    public AlleleFrequencyAnalyzer()
73      : base() {
74      Parameters.Add(new LookupParameter<BoolValue>("Maximization", "True if the problem is a maximization problem."));
75      Parameters.Add(new ScopeTreeLookupParameter<T>("Solution", "The solutions whose alleles should be analyzed."));
76      Parameters.Add(new ScopeTreeLookupParameter<DoubleValue>("Quality", "The qualities of the solutions which should be analyzed."));
77      Parameters.Add(new LookupParameter<T>("BestKnownSolution", "The best known solution."));
78      Parameters.Add(new ValueLookupParameter<ResultCollection>("Results", "The result collection where the allele frequency analysis results should be stored."));
79      Parameters.Add(new ValueParameter<BoolValue>("StoreHistory", "True if the history of the allele frequency analysis should be stored.", new BoolValue(false)));
80      Parameters.Add(new ValueParameter<IntValue>("UpdateInterval", "The interval in which the allele frequency analysis should be applied.", new IntValue(1)));
81      Parameters.Add(new LookupParameter<IntValue>("UpdateCounter", "The value which counts how many times the operator was called since the last update.", "AlleleFrequencyAnalyzerUpdateCounter"));
82
83      MaximizationParameter.Hidden = true;
84      SolutionParameter.Hidden = true;
85      QualityParameter.Hidden = true;
86      BestKnownSolutionParameter.Hidden = true;
87      ResultsParameter.Hidden = true;
88      UpdateCounterParameter.Hidden = true;
89    }
90
91    #region Equality Comparers
92    private class AlleleIdEqualityComparer : IEqualityComparer<Allele> {
93      public bool Equals(Allele x, Allele y) {
94        return x.Id == y.Id;
95      }
96      public int GetHashCode(Allele obj) {
97        return obj.Id.GetHashCode();
98      }
99    }
100    private class AlleleFrequencyIdEqualityComparer : IEqualityComparer<AlleleFrequency> {
101      public bool Equals(AlleleFrequency x, AlleleFrequency y) {
102        return x.Id == y.Id;
103      }
104      public int GetHashCode(AlleleFrequency obj) {
105        return obj.Id.GetHashCode();
106      }
107    }
108    #endregion
109
110    public override IOperation Apply() {
111      int updateInterval = UpdateIntervalParameter.Value.Value;
112      IntValue updateCounter = UpdateCounterParameter.ActualValue;
113      if (updateCounter == null) {
114        updateCounter = new IntValue(updateInterval);
115        UpdateCounterParameter.ActualValue = updateCounter;
116      } else updateCounter.Value++;
117
118      if (updateCounter.Value == updateInterval) {
119        updateCounter.Value = 0;
120
121        bool max = MaximizationParameter.ActualValue.Value;
122        ItemArray<T> solutions = SolutionParameter.ActualValue;
123        double[] qualities = QualityParameter.ActualValue.Select(x => x.Value).ToArray();
124        T bestKnownSolution = BestKnownSolutionParameter.ActualValue;
125        bool storeHistory = StoreHistoryParameter.Value.Value;
126
127        // calculate index of current best solution
128        int bestIndex = -1;
129        if (!max) {
130          bestIndex = qualities
131            .Select((x, i) => new { Index = i, Value = x })
132            .OrderBy(x => x.Value)
133            .First().Index;
134        } else {
135          bestIndex = qualities
136            .Select((x, i) => new { Index = i, Value = x })
137            .OrderByDescending(x => x.Value)
138            .First().Index;
139        }
140
141        // calculate allels of current best and (if available) best known solution
142        Allele[] bestAlleles = CalculateAlleles(solutions[bestIndex]);
143        Allele[] bestKnownAlleles = null;
144        if (bestKnownSolution != null)
145          bestKnownAlleles = CalculateAlleles(bestKnownSolution);
146
147        // calculate allele frequencies
148        var frequencies = solutions.SelectMany((s, index) => CalculateAlleles(s).Select(a => new { Allele = a, Quality = qualities[index] })).
149                          GroupBy(x => x.Allele.Id).
150                          Select(x => new AlleleFrequency(x.Key,
151                                                          x.Count() / ((double)solutions.Length),
152                                                          x.Average(a => a.Allele.Impact),
153                                                          x.Average(a => a.Quality),
154                                                          bestKnownAlleles == null ? false : bestKnownAlleles.Any(a => a.Id == x.Key),
155                                                          bestAlleles.Any(a => a.Id == x.Key)));
156
157        // calculate dummy allele frequencies of alleles of best known solution which did not occur
158        if (bestKnownAlleles != null) {
159          var bestKnownFrequencies = bestKnownAlleles.Select(x => new AlleleFrequency(x.Id, 0, x.Impact, 0, true, false)).Except(frequencies, new AlleleFrequencyIdEqualityComparer());
160          frequencies = frequencies.Concat(bestKnownFrequencies);
161        }
162
163        // fetch results collection
164        ResultCollection results;
165        if (!ResultsParameter.ActualValue.ContainsKey(Name + " Results")) {
166          results = new ResultCollection();
167          ResultsParameter.ActualValue.Add(new Result(Name + " Results", results));
168        } else {
169          results = (ResultCollection)ResultsParameter.ActualValue[Name + " Results"].Value;
170        }
171
172        // store allele frequencies
173        AlleleFrequencyCollection frequenciesCollection = new AlleleFrequencyCollection(frequencies);
174        if (!results.ContainsKey("Allele Frequencies"))
175          results.Add(new Result("Allele Frequencies", frequenciesCollection));
176        else
177          results["Allele Frequencies"].Value = frequenciesCollection;
178
179        // store allele frequencies history
180        if (storeHistory) {
181          if (!results.ContainsKey("Allele Frequencies History")) {
182            AlleleFrequencyCollectionHistory history = new AlleleFrequencyCollectionHistory();
183            history.Add(frequenciesCollection);
184            results.Add(new Result("Allele Frequencies History", history));
185          } else {
186            ((AlleleFrequencyCollectionHistory)results["Allele Frequencies History"].Value).Add(frequenciesCollection);
187          }
188        }
189
190        // store alleles data table
191        DataTable allelesTable;
192        if (!results.ContainsKey("Alleles")) {
193          allelesTable = new DataTable("Alleles");
194          allelesTable.VisualProperties.XAxisTitle = "Iteration";
195          allelesTable.VisualProperties.YAxisTitle = "Number of Alleles";
196          allelesTable.VisualProperties.SecondYAxisTitle = "Number of Alleles";
197
198          allelesTable.Rows.Add(new DataRow("Unique Alleles"));
199          allelesTable.Rows["Unique Alleles"].VisualProperties.StartIndexZero = true;
200
201          allelesTable.Rows.Add(new DataRow("Unique Alleles of Best Known Solution", null));
202          allelesTable.Rows["Unique Alleles of Best Known Solution"].VisualProperties.SecondYAxis = true;
203          allelesTable.Rows["Unique Alleles of Best Known Solution"].VisualProperties.StartIndexZero = true;
204
205          allelesTable.Rows.Add(new DataRow("Fixed Alleles", null));
206          allelesTable.Rows["Fixed Alleles"].VisualProperties.SecondYAxis = true;
207          allelesTable.Rows["Fixed Alleles"].VisualProperties.StartIndexZero = true;
208
209          allelesTable.Rows.Add(new DataRow("Fixed Alleles of Best Known Solution", null));
210          allelesTable.Rows["Fixed Alleles of Best Known Solution"].VisualProperties.SecondYAxis = true;
211          allelesTable.Rows["Fixed Alleles of Best Known Solution"].VisualProperties.StartIndexZero = true;
212
213          allelesTable.Rows.Add(new DataRow("Lost Alleles of Best Known Solution", null));
214          allelesTable.Rows["Lost Alleles of Best Known Solution"].VisualProperties.SecondYAxis = true;
215          allelesTable.Rows["Lost Alleles of Best Known Solution"].VisualProperties.StartIndexZero = true;
216
217          results.Add(new Result("Alleles", allelesTable));
218        } else {
219          allelesTable = (DataTable)results["Alleles"].Value;
220        }
221
222        int fixedAllelesCount = frequenciesCollection.Where(x => x.Frequency == 1).Count();
223        var relevantAlleles = frequenciesCollection.Where(x => x.ContainedInBestKnownSolution);
224        int relevantAllelesCount = relevantAlleles.Count();
225        int fixedRelevantAllelesCount = relevantAlleles.Where(x => x.Frequency == 1).Count();
226        int lostRelevantAllelesCount = relevantAlleles.Where(x => x.Frequency == 0).Count();
227        int uniqueRelevantAllelesCount = relevantAllelesCount - lostRelevantAllelesCount;
228        allelesTable.Rows["Unique Alleles"].Values.Add(frequenciesCollection.Count);
229        allelesTable.Rows["Unique Alleles of Best Known Solution"].Values.Add(uniqueRelevantAllelesCount);
230        allelesTable.Rows["Fixed Alleles"].Values.Add(fixedAllelesCount);
231        allelesTable.Rows["Fixed Alleles of Best Known Solution"].Values.Add(fixedRelevantAllelesCount);
232        allelesTable.Rows["Lost Alleles of Best Known Solution"].Values.Add(lostRelevantAllelesCount);
233
234        // store alleles values
235        if (!results.ContainsKey("Unique Alleles"))
236          results.Add(new Result("Unique Alleles", new DoubleValue(frequenciesCollection.Count)));
237        else
238          ((DoubleValue)results["Unique Alleles"].Value).Value = frequenciesCollection.Count;
239
240        if (!results.ContainsKey("Unique Alleles of Best Known Solution"))
241          results.Add(new Result("Unique Alleles of Best Known Solution", new DoubleValue(uniqueRelevantAllelesCount)));
242        else
243          ((DoubleValue)results["Unique Alleles of Best Known Solution"].Value).Value = uniqueRelevantAllelesCount;
244
245        if (!results.ContainsKey("Fixed Alleles"))
246          results.Add(new Result("Fixed Alleles", new DoubleValue(fixedAllelesCount)));
247        else
248          ((DoubleValue)results["Fixed Alleles"].Value).Value = fixedAllelesCount;
249
250        if (!results.ContainsKey("Fixed Alleles of Best Known Solution"))
251          results.Add(new Result("Fixed Alleles of Best Known Solution", new DoubleValue(fixedRelevantAllelesCount)));
252        else
253          ((DoubleValue)results["Fixed Alleles of Best Known Solution"].Value).Value = fixedRelevantAllelesCount;
254
255        if (!results.ContainsKey("Lost Alleles of Best Known Solution"))
256          results.Add(new Result("Lost Alleles of Best Known Solution", new DoubleValue(lostRelevantAllelesCount)));
257        else
258          ((DoubleValue)results["Lost Alleles of Best Known Solution"].Value).Value = lostRelevantAllelesCount;
259
260        // calculate contained alleles of best known solution and relative quality
261        if (bestKnownAlleles != null) {
262          double qualityRange = Math.Abs(qualities.Max() - qualities.Min());
263          var points = solutions.Select((s, index) => new Point2D<double>(CalculateAlleles(s).Intersect(bestKnownAlleles, new AlleleIdEqualityComparer()).Count(),
264                                                                          Math.Abs(qualities[index] - qualities[bestIndex]) / qualityRange));
265          var avgContainedReleventAlleles = points.Select(x => x.X).Average();
266
267          var plot = new ScatterPlot("Contained Alleles of Best Known Solution and Relative Solution Qualtiy", null);
268          plot.VisualProperties.XAxisTitle = "Contained Alleles of Best Known Solution";
269          plot.VisualProperties.YAxisTitle = "Relative Solution Quality";
270          plot.VisualProperties.XAxisMinimumAuto = false;
271          plot.VisualProperties.XAxisMinimumFixedValue = 0.0;
272          plot.VisualProperties.XAxisMaximumAuto = false;
273          plot.VisualProperties.XAxisMaximumFixedValue = bestKnownAlleles.Length;
274          plot.VisualProperties.YAxisMinimumAuto = false;
275          plot.VisualProperties.YAxisMinimumFixedValue = 0.0;
276          plot.VisualProperties.YAxisMaximumAuto = false;
277          plot.VisualProperties.YAxisMaximumFixedValue = 1.0;
278          var row = new ScatterPlotDataRow("Solutions of Current Generation", null, points);
279          row.VisualProperties.PointStyle = ScatterPlotDataRowVisualProperties.ScatterPlotDataRowPointStyle.Circle;
280          row.VisualProperties.PointSize = 5;
281          plot.Rows.Add(row);
282
283          if (!results.ContainsKey("Scatter Plot"))
284            results.Add(new Result("Scatter Plot", plot));
285          else
286            results["Scatter Plot"].Value = plot;
287          if (storeHistory) {
288            if (!results.ContainsKey("Scatter Plot History")) {
289              results.Add(new Result("Scatter Plot History", new ScatterPlotHistory()));
290            }
291            ((ScatterPlotHistory)results["Scatter Plot History"].Value).Add(plot);
292          }
293
294          if (!allelesTable.Rows.ContainsKey("Average Contained Alleles of Best Known Solution")) {
295            allelesTable.Rows.Add(new DataRow("Average Contained Alleles of Best Known Solution", null));
296            allelesTable.Rows["Average Contained Alleles of Best Known Solution"].VisualProperties.SecondYAxis = true;
297            allelesTable.Rows["Average Contained Alleles of Best Known Solution"].VisualProperties.StartIndexZero = true;
298          }
299          allelesTable.Rows["Average Contained Alleles of Best Known Solution"].Values.Add(avgContainedReleventAlleles);
300
301          if (!results.ContainsKey("Average Contained Alleles of Best Known Solution"))
302            results.Add(new Result("Average Contained Alleles of Best Known Solution", new DoubleValue(avgContainedReleventAlleles)));
303          else
304            ((DoubleValue)results["Average Contained Alleles of Best Known Solution"].Value).Value = avgContainedReleventAlleles;
305        }
306      }
307      return base.Apply();
308    }
309
310    protected abstract Allele[] CalculateAlleles(T solution);
311  }
312}
Note: See TracBrowser for help on using the repository browser.