Changeset 12988 for branches/HeuristicLab.EvolutionTracking/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Tracking/SchemaDiversification/DiversificationStatisticsOperator.cs

Timestamp:

10/09/15 00:16:03 (9 years ago)

Author:

bburlacu

Message:

#1772: Performance improvement changes

• QueryMatch.cs: eliminate unnecessary ToList() call and expensive GetBranchLevel calls
• Diversification: eliminated creation of shallow copies of the individual subscopes as it was either too slow (due to events being registered/deregistered when variables are added to the scope) or too leaky (if attempting to clear the scopes without clearing the variables then the code is leaking EventHandlers)
• Aggregated diversification statistics separately with the help of some parameters set up in the SchemaCreator and SchemaEvaluator
• Made code in the UpdateEstimatedValuesOperator perform exactly as in the evaluator (updating quality and estimated values)
• Removed no longer needed SchemaCleanupOperator
• Do not evaluate intermediate vertices in the genealogy analyzer if the TrimOlderGenerations flag is activated

New functionality:

• parameter to control the fraction of the population to be considered by the diversification strategy
• parameter to control whether individuals may be matched by any schema and mutated only once (exclusive matching)
• parameter to control whether linear scaling should be applied to the estimated values used for the calculation of phenotypic similarity (default: yes)

File:

• branches/HeuristicLab.EvolutionTracking/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Tracking/SchemaDiversification/DiversificationStatisticsOperator.cs (moved) (moved from branches/HeuristicLab.EvolutionTracking/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Tracking/SchemaDiversification/SchemaCleanupOperator.cs) (2 diffs)

branches/HeuristicLab.EvolutionTracking/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Tracking/SchemaDiversification/DiversificationStatisticsOperator.cs

@@ -20 +20 @@
 #endregion
 
-using System.Linq;
+using HeuristicLab.Analysis;
 using HeuristicLab.Common;
 using HeuristicLab.Core;
+using HeuristicLab.Data;
 using HeuristicLab.Operators;
+using HeuristicLab.Optimization;
+using HeuristicLab.Parameters;
 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
 
@@ -29 +32 @@
   [Item("SchemaCleanupOperator", "Operator which removes the schemas from the global scope after they have been evaluated.")]
   [StorableClass]
-  public class SchemaCleanupOperator : SingleSuccessorOperator {
-    public SchemaCleanupOperator() { }
+  public class DiversificationStatisticsOperator : SingleSuccessorOperator {
+    private const string NumberOfChangedTreesParameterName = "NumberOfChangedTrees";
+    private const string NumberOfSchemasParameterName = "NumberOfSchemas";
+    private const string AverageSchemaLengthParameterName = "AverageSchemaLength";
+    private const string ResultCollectionParameterName = "Results";
 
-    protected SchemaCleanupOperator(SchemaCleanupOperator original, Cloner cloner) : base(original, cloner) { }
+    public ILookupParameter<IntValue> NumberOfChangedTreesParameter {
+      get { return (ILookupParameter<IntValue>)Parameters[NumberOfChangedTreesParameterName]; }
+    }
+    public ILookupParameter<IntValue> NumberOfSchemasParameter {
+      get { return (ILookupParameter<IntValue>)Parameters[NumberOfSchemasParameterName]; }
+    }
+    public ILookupParameter<DoubleValue> AverageSchemaLengthParameter {
+      get { return (ILookupParameter<DoubleValue>)Parameters[AverageSchemaLengthParameterName]; }
+    }
+    public ILookupParameter<ResultCollection> ResultCollectionParameter {
+      get { return (ILookupParameter<ResultCollection>)Parameters[ResultCollectionParameterName]; }
+    }
 
-    public override IDeepCloneable Clone(Cloner cloner) { return new SchemaCleanupOperator(this, cloner); }
+    public DiversificationStatisticsOperator() {
+      Parameters.Add(new LookupParameter<IntValue>(NumberOfChangedTreesParameterName));
+      Parameters.Add(new LookupParameter<IntValue>(NumberOfSchemasParameterName));
+      Parameters.Add(new LookupParameter<DoubleValue>(AverageSchemaLengthParameterName));
+      Parameters.Add(new LookupParameter<ResultCollection>(ResultCollectionParameterName));
+    }
+
+    protected DiversificationStatisticsOperator(DiversificationStatisticsOperator original, Cloner cloner) : base(original, cloner) { }
+
+    public override IDeepCloneable Clone(Cloner cloner) { return new DiversificationStatisticsOperator(this, cloner); }
 
     [StorableConstructor]
-    protected SchemaCleanupOperator(bool deserializing) : base(deserializing) { }
+    protected DiversificationStatisticsOperator(bool deserializing) : base(deserializing) { }
 
     public override IOperation Apply() {
-      foreach (var scope in ExecutionContext.Scope.SubScopes.Where(x => x.Name.Equals("Schema"))) {
-        foreach (var subscope in scope.SubScopes) { subscope.Variables.Clear(); }
-        scope.SubScopes.Clear();
-        scope.Variables.Clear();
+
+      var results = ResultCollectionParameter.ActualValue;
+      DataTable table;
+      if (!results.ContainsKey("NumberOfChangedTrees")) {
+        table = new DataTable();
+        results.Add(new Result("NumberOfChangedTrees", table));
+        var row = new DataRow("Changed trees");
+        table.Rows.Add(row);
       }
-      ExecutionContext.Scope.SubScopes.RemoveAll(x => x.Name.Equals("Schema"));
+      if (!results.ContainsKey("AverageSchemaLength")) {
+        table = new DataTable();
+        results.Add(new Result("AverageSchemaLength", table));
+        var row = new DataRow("Average schema length");
+        table.Rows.Add(row);
+      }
+      if (!results.ContainsKey("NumberOfSchemas")) {
+        table = new DataTable();
+        results.Add(new Result("NumberOfSchemas", table));
+        var row = new DataRow("Number of schemas");
+        table.Rows.Add(row);
+      }
+      ((DataTable)results["NumberOfChangedTrees"].Value).Rows["Changed trees"].Values.Add(NumberOfChangedTreesParameter.ActualValue.Value);
+      ((DataTable)results["AverageSchemaLength"].Value).Rows["Average schema length"].Values.Add(AverageSchemaLengthParameter.ActualValue.Value);
+      ((DataTable)results["NumberOfSchemas"].Value).Rows["Number of schemas"].Values.Add(NumberOfSchemasParameter.ActualValue.Value);
+
       return base.Apply();
     }