Changeset 12378 for branches/GBT/HeuristicLab.Problems.DataAnalysis.Symbolic

Timestamp:

05/04/15 13:08:49 (10 years ago)

Author:

gkronber

Message:

Merged revision(s) 12333-12365 from trunk/sources:
#2373: Corrected typo in RandomBinaryVectorCreator by implementing an after-deserialization-hook.

........
#2359: Refactored pruning operators and analyzers.

........
#2359: Removed commented code from pruning analyzer.

........
#2378: Vertex.cs: Fixed bug in Label setter.
........
#2359: The changes in r12358 look fine to me. Added total number of pruned nodes in the analyzer's data table. Removed unused parameter names in the SymbolicDataAnalysisSingleObjectivePruningAnalyzer.
........
#2345: Fixed x-axis maximum in error characteristics curve.

........

Location:

branches/GBT

Files:

• . (modified) (1 prop)
• HeuristicLab.Problems.DataAnalysis.Symbolic (modified) (1 prop)
• HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Analyzers/SymbolicDataAnalysisSingleObjectivePruningAnalyzer.cs (modified) (10 diffs)
• HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/SymbolicDataAnalysisExpressionPruningOperator.cs (modified) (8 diffs)

branches/GBT/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Analyzers/SymbolicDataAnalysisSingleObjectivePruningAnalyzer.cs

@@ -45 +45 @@
     private const string TotalNumberOfPrunedSubtreesParameterName = "Number of pruned subtrees";
     private const string TotalNumberOfPrunedTreesParameterName = "Number of pruned trees";
+    private const string TotalNumberOfPrunedNodesParameterName = "Number of pruned nodes";
     private const string RandomParameterName = "Random";
-    private const string PruneOnlyZeroImpactNodesParameterName = "PruneOnlyZeroImpactNodes";
-    private const string NodeImpactThresholdParameterName = "ImpactThreshold";
-    private const string PruningOperatorParameterName = "PruningOperator";
     private const string ResultsParameterName = "Results";
     private const string PopulationSizeParameterName = "PopulationSize";
@@ -54 +52 @@
 
     #region private members
+    private DataReducer prunedNodesReducer;
     private DataReducer prunedSubtreesReducer;
     private DataReducer prunedTreesReducer;
@@ -61 +60 @@
 
     #region parameter properties
-    public IValueParameter<SymbolicDataAnalysisExpressionPruningOperator> PruningOperatorParameter {
-      get { return (IValueParameter<SymbolicDataAnalysisExpressionPruningOperator>)Parameters[PruningOperatorParameterName]; }
-    }
-    public IFixedValueParameter<BoolValue> PruneOnlyZeroImpactNodesParameter {
-      get { return (IFixedValueParameter<BoolValue>)Parameters[PruneOnlyZeroImpactNodesParameterName]; }
-    }
-    public IFixedValueParameter<DoubleValue> NodeImpactThresholdParameter {
-      get { return (IFixedValueParameter<DoubleValue>)Parameters[NodeImpactThresholdParameterName]; }
-    }
     public ILookupParameter<IRandom> RandomParameter {
       get { return (ILookupParameter<IRandom>)Parameters[RandomParameterName]; }
@@ -91 +81 @@
 
     #region properties
-    protected SymbolicDataAnalysisExpressionPruningOperator PruningOperator { get { return PruningOperatorParameter.Value; } }
+    protected abstract SymbolicDataAnalysisExpressionPruningOperator PruningOperator { get; }
     protected int UpdateInterval { get { return UpdateIntervalParameter.Value.Value; } }
 
@@ -112 +102 @@
       get { return PruningProbabilityParameter.Value.Value; }
       set { PruningProbabilityParameter.Value.Value = value; }
-    }
-
-    protected bool PruneOnlyZeroImpactNodes {
-      get { return PruneOnlyZeroImpactNodesParameter.Value.Value; }
-      set { PruneOnlyZeroImpactNodesParameter.Value.Value = value; }
-    }
-    protected double NodeImpactThreshold {
-      get { return NodeImpactThresholdParameter.Value.Value; }
-      set { NodeImpactThresholdParameter.Value.Value = value; }
     }
     #endregion
@@ -140 +121 @@
     protected SymbolicDataAnalysisSingleObjectivePruningAnalyzer(SymbolicDataAnalysisSingleObjectivePruningAnalyzer original, Cloner cloner)
       : base(original, cloner) {
+      if (original.prunedNodesReducer != null)
+        this.prunedNodesReducer = (DataReducer)original.prunedNodesReducer.Clone();
       if (original.prunedSubtreesReducer != null)
         this.prunedSubtreesReducer = (DataReducer)original.prunedSubtreesReducer.Clone();
@@ -197 +180 @@
       Parameters.Add(new LookupParameter<IRandom>(RandomParameterName, "The random number generator."));
       Parameters.Add(new LookupParameter<IDataAnalysisProblemData>(ProblemDataParameterName, "The problem data."));
-      Parameters.Add(new FixedValueParameter<DoubleValue>(NodeImpactThresholdParameterName, "The impact threshold  below which an individual should be pruned.", new DoubleValue(0.0)));
-      Parameters.Add(new FixedValueParameter<BoolValue>(PruneOnlyZeroImpactNodesParameterName, "Switch to determine of only zero impact individuals should be pruned.", new BoolValue(false)));
       Parameters.Add(new LookupParameter<IntValue>(PopulationSizeParameterName, "The population of individuals."));
       #endregion
@@ -250 +231 @@
       UpdateCounter = 0;
 
-      if (prunedSubtreesReducer == null || prunedTreesReducer == null || valuesCollector == null || resultsCollector == null) { InitializeOperators(); }
+      if (prunedNodesReducer == null || prunedSubtreesReducer == null || prunedTreesReducer == null || valuesCollector == null || resultsCollector == null) { InitializeOperators(); }
 
       var prune = CreatePruningOperation();
+      var reducePrunedNodes = ExecutionContext.CreateChildOperation(prunedNodesReducer);
       var reducePrunedSubtrees = ExecutionContext.CreateChildOperation(prunedSubtreesReducer);
       var reducePrunedTrees = ExecutionContext.CreateChildOperation(prunedTreesReducer);
@@ -258 +240 @@
       var collectResults = ExecutionContext.CreateChildOperation(resultsCollector);
 
-      return new OperationCollection { prune, reducePrunedSubtrees, reducePrunedTrees, collectValues, collectResults, base.Apply() };
+      return new OperationCollection { prune, reducePrunedNodes, reducePrunedSubtrees, reducePrunedTrees, collectValues, collectResults, base.Apply() };
     }
 
     private void InitializeOperators() {
+      prunedNodesReducer = new DataReducer();
+      prunedNodesReducer.ParameterToReduce.ActualName = PruningOperator.PrunedNodesParameter.ActualName;
+      prunedNodesReducer.ReductionOperation.Value = new ReductionOperation(ReductionOperations.Sum); // sum all the pruned subtrees parameter values
+      prunedNodesReducer.TargetOperation.Value = new ReductionOperation(ReductionOperations.Assign); // asign the sum to the target parameter
+      prunedNodesReducer.TargetParameter.ActualName = TotalNumberOfPrunedNodesParameterName;
+
       prunedSubtreesReducer = new DataReducer();
       prunedSubtreesReducer.ParameterToReduce.ActualName = PruningOperator.PrunedSubtreesParameter.ActualName;
@@ -275 +263 @@
 
       valuesCollector = new DataTableValuesCollector();
+      valuesCollector.CollectedValues.Add(new LookupParameter<IntValue>(TotalNumberOfPrunedNodesParameterName));
       valuesCollector.CollectedValues.Add(new LookupParameter<IntValue>(TotalNumberOfPrunedSubtreesParameterName));
       valuesCollector.CollectedValues.Add(new LookupParameter<IntValue>(TotalNumberOfPrunedTreesParameterName));

branches/GBT/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/SymbolicDataAnalysisExpressionPruningOperator.cs

@@ -34 +34 @@
   [StorableClass]
   [Item("SymbolicExpressionTreePruningOperator", "An operator that replaces introns with constant values in a symbolic expression tree.")]
-  public abstract class SymbolicDataAnalysisExpressionPruningOperator : SingleSuccessorOperator {
+  public abstract class SymbolicDataAnalysisExpressionPruningOperator : SingleSuccessorOperator, ISymbolicExpressionTreeOperator {
     #region parameter names
     private const string ProblemDataParameterName = "ProblemData";
@@ -41 +41 @@
     private const string PrunedSubtreesParameterName = "PrunedSubtrees";
     private const string PrunedTreesParameterName = "PrunedTrees";
+    private const string PrunedNodesParameterName = "PrunedNodes";
     private const string FitnessCalculationPartitionParameterName = "FitnessCalculationPartition";
     private const string NodeImpactThresholdParameterName = "ImpactThreshold";
@@ -72 +73 @@
       get { return (ILookupParameter<IntValue>)Parameters[PrunedTreesParameterName]; }
     }
+    public ILookupParameter<IntValue> PrunedNodesParameter {
+      get { return (ILookupParameter<IntValue>)Parameters[PrunedNodesParameterName]; }
+    }
     public IFixedValueParameter<DoubleValue> NodeImpactThresholdParameter {
       get { return (IFixedValueParameter<DoubleValue>)Parameters[NodeImpactThresholdParameterName]; }
@@ -87 +91 @@
 
     #region properties
-    protected IDataAnalysisProblemData ProblemData { get { return ProblemDataParameter.ActualValue; } }
-    protected ISymbolicDataAnalysisSolutionImpactValuesCalculator ImpactValuesCalculator { get { return ImpactValuesCalculatorParameter.Value; } }
-    protected IntRange FitnessCalculationPartition { get { return FitnessCalculationPartitionParameter.ActualValue; } }
-    protected bool PruneOnlyZeroImpactNodes {
+    public ISymbolicDataAnalysisSolutionImpactValuesCalculator ImpactValuesCalculator {
+      get { return ImpactValuesCalculatorParameter.Value; }
+      set { ImpactValuesCalculatorParameter.Value = value; }
+    }
+    public bool PruneOnlyZeroImpactNodes {
       get { return PruneOnlyZeroImpactNodesParameter.Value.Value; }
       set { PruneOnlyZeroImpactNodesParameter.Value.Value = value; }
     }
-    protected double NodeImpactThreshold {
+    public double NodeImpactThreshold {
       get { return NodeImpactThresholdParameter.Value.Value; }
       set { NodeImpactThresholdParameter.Value.Value = value; }
     }
-    protected ISymbolicExpressionTree SymbolicExpressionTree { get { return SymbolicExpressionTreeParameter.ActualValue; } }
-    protected DoubleValue Quality { get { return QualityParameter.ActualValue; } }
-    protected DoubleLimit EstimationLimits { get { return EstimationLimitsParameter.ActualValue; } }
-    protected ISymbolicDataAnalysisExpressionTreeInterpreter Interpreter { get { return InterpreterParameter.ActualValue; } }
     #endregion
 
@@ -114 +115 @@
       Parameters.Add(new LookupParameter<ISymbolicDataAnalysisModel>(SymbolicDataAnalysisModelParameterName));
       Parameters.Add(new LookupParameter<IntRange>(FitnessCalculationPartitionParameterName));
+      Parameters.Add(new LookupParameter<IntValue>(PrunedNodesParameterName, "A counter of how many nodes were pruned."));
       Parameters.Add(new LookupParameter<IntValue>(PrunedSubtreesParameterName, "A counter of how many subtrees were replaced."));
       Parameters.Add(new LookupParameter<IntValue>(PrunedTreesParameterName, "A counter of how many trees were pruned."));
@@ -131 +133 @@
 
     public override IOperation Apply() {
-      var model = CreateModel(SymbolicExpressionTree, Interpreter, ProblemData, EstimationLimits);
-      var nodes = SymbolicExpressionTree.Root.GetSubtree(0).GetSubtree(0).IterateNodesPrefix().ToList();
-      var rows = Enumerable.Range(FitnessCalculationPartition.Start, FitnessCalculationPartition.Size);
+      var tree = SymbolicExpressionTreeParameter.ActualValue;
+      var problemData = ProblemDataParameter.ActualValue;
+      var fitnessCalculationPartition = FitnessCalculationPartitionParameter.ActualValue;
+      var estimationLimits = EstimationLimitsParameter.ActualValue;
+      var interpreter = InterpreterParameter.ActualValue;
+
+      var model = CreateModel(tree, interpreter, problemData, estimationLimits);
+      var nodes = tree.Root.GetSubtree(0).GetSubtree(0).IterateNodesPrefix().ToList();
+      var rows = Enumerable.Range(fitnessCalculationPartition.Start, fitnessCalculationPartition.Size);
       var prunedSubtrees = 0;
       var prunedTrees = 0;
+      var prunedNodes = 0;
 
       double quality = Evaluate(model);
@@ -144 +153 @@
 
         double impactValue, replacementValue;
-        ImpactValuesCalculator.CalculateImpactAndReplacementValues(model, node, ProblemData, rows, out impactValue, out replacementValue, quality);
+        ImpactValuesCalculator.CalculateImpactAndReplacementValues(model, node, problemData, rows, out impactValue, out replacementValue, quality);
 
-        if (PruneOnlyZeroImpactNodes) {
-          if (!impactValue.IsAlmost(0.0)) continue;
-        } else if (NodeImpactThreshold < impactValue) {
-          continue;
-        }
+        if (PruneOnlyZeroImpactNodes && !impactValue.IsAlmost(0.0)) continue;
+        if (!PruneOnlyZeroImpactNodes && impactValue > NodeImpactThreshold) continue;
 
         var constantNode = (ConstantTreeNode)node.Grammar.GetSymbol("Constant").CreateTreeNode();
         constantNode.Value = replacementValue;
 
+        var length = node.GetLength();
         ReplaceWithConstant(node, constantNode);
-        i += node.GetLength() - 1; // skip subtrees under the node that was folded
+        i += length - 1; // skip subtrees under the node that was folded
 
         quality -= impactValue;
-
         prunedSubtrees++;
+        prunedNodes += length;
       }
 
@@ -166 +173 @@
       PrunedSubtreesParameter.ActualValue = new IntValue(prunedSubtrees);
       PrunedTreesParameter.ActualValue = new IntValue(prunedTrees);
+      PrunedNodesParameter.ActualValue = new IntValue(prunedNodes);
+      QualityParameter.ActualValue.Value = quality;
 
       return base.Apply();
-    }
-
-    public ISymbolicExpressionTree Prune(ISymbolicExpressionTree tree, ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, IDataAnalysisProblemData problemData, DoubleLimit estimationLimits) {
-      var model = CreateModel((ISymbolicExpressionTree)tree.Clone(), Interpreter, ProblemData, EstimationLimits);
-      var nodes = SymbolicExpressionTree.Root.GetSubtree(0).GetSubtree(0).IterateNodesPrefix().ToList();
-      var rows = Enumerable.Range(FitnessCalculationPartition.Start, FitnessCalculationPartition.Size);
-
-      double quality = Evaluate(model);
-
-      for (int i = 0; i < nodes.Count; ++i) {
-        var node = nodes[i];
-        if (node is ConstantTreeNode) continue;
-
-        double impactValue, replacementValue;
-        ImpactValuesCalculator.CalculateImpactAndReplacementValues(model, node, ProblemData, rows, out impactValue, out replacementValue, quality);
-
-        if (PruneOnlyZeroImpactNodes) {
-          if (!impactValue.IsAlmost(0.0)) continue;
-        } else if (NodeImpactThreshold < impactValue) {
-          continue;
-        }
-
-        var constantNode = (ConstantTreeNode)node.Grammar.GetSymbol("Constant").CreateTreeNode();
-        constantNode.Value = replacementValue;
-
-        ReplaceWithConstant(node, constantNode);
-        i += node.GetLength() - 1; // skip subtrees under the node that was folded
-
-        quality -= impactValue;
-      }
-      return model.SymbolicExpressionTree;
     }