Changeset 15371 for trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Classification/3.4/SymbolicClassificationSolutionImpactValuesCalculator.cs

Timestamp:

09/19/17 10:49:02 (7 years ago)

Author:

pfleck

Message:

#1666: Merged into trunk

Location:

trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Classification

Files:

• . (modified) (1 prop)
• 3.4/SymbolicClassificationSolutionImpactValuesCalculator.cs (modified) (2 diffs)

trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Classification/3.4/SymbolicClassificationSolutionImpactValuesCalculator.cs

@@ -20 +20 @@
 #endregion
 
-using System;
 using System.Collections.Generic;
 using HeuristicLab.Common;
 using HeuristicLab.Core;
-using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
 
@@ -40 +38 @@
     protected SymbolicClassificationSolutionImpactValuesCalculator(bool deserializing) : base(deserializing) { }
 
-    public override void CalculateImpactAndReplacementValues(ISymbolicDataAnalysisModel model,
-      ISymbolicExpressionTreeNode node,
-      IDataAnalysisProblemData problemData, IEnumerable<int> rows, out double impactValue, out double replacementValue,
-      out double newQualityForImpactsCalculation,
-      double qualityForImpactsCalculation = Double.NaN) {
+    protected override double CalculateQualityForImpacts(ISymbolicDataAnalysisModel model, IDataAnalysisProblemData problemData, IEnumerable<int> rows) {
       var classificationModel = (ISymbolicClassificationModel)model;
       var classificationProblemData = (IClassificationProblemData)problemData;
-
-      if (double.IsNaN(qualityForImpactsCalculation))
-        qualityForImpactsCalculation = CalculateQualityForImpacts(classificationModel, classificationProblemData, rows);
-
-
-      var cloner = new Cloner();
-      var tempModel = cloner.Clone(classificationModel);
-      var tempModelNode = (ISymbolicExpressionTreeNode)cloner.GetClone(node);
-
-      var tempModelParentNode = tempModelNode.Parent;
-      int i = tempModelParentNode.IndexOfSubtree(tempModelNode);
-      double bestReplacementValue = 0.0;
-      double bestImpactValue = double.PositiveInfinity;
-      newQualityForImpactsCalculation = qualityForImpactsCalculation; // initialize
-      // try the potentially reasonable replacement values and use the best one
-      foreach (var repValue in CalculateReplacementValues(node, classificationModel.SymbolicExpressionTree, classificationModel.Interpreter, classificationProblemData.Dataset, classificationProblemData.TrainingIndices)) {
-        tempModelParentNode.RemoveSubtree(i);
-
-        var constantNode = new ConstantTreeNode(new Constant()) { Value = repValue };
-        tempModelParentNode.InsertSubtree(i, constantNode);
-
-        var dataset = classificationProblemData.Dataset;
-        var targetClassValues = dataset.GetDoubleValues(classificationProblemData.TargetVariable, rows);
-        var estimatedClassValues = tempModel.GetEstimatedClassValues(dataset, rows);
-        OnlineCalculatorError errorState;
-        newQualityForImpactsCalculation = OnlineAccuracyCalculator.Calculate(targetClassValues, estimatedClassValues,
-          out errorState);
-        if (errorState != OnlineCalculatorError.None) newQualityForImpactsCalculation = 0.0;
-
-        impactValue = qualityForImpactsCalculation - newQualityForImpactsCalculation;
-
-        if (impactValue < bestImpactValue) {
-          bestImpactValue = impactValue;
-          bestReplacementValue = repValue;
-        }
-      }
-      replacementValue = bestReplacementValue;
-      impactValue = bestImpactValue;
-    }
-
-    public static double CalculateQualityForImpacts(ISymbolicClassificationModel model, IClassificationProblemData problemData, IEnumerable<int> rows) {
       OnlineCalculatorError errorState;
       var dataset = problemData.Dataset;
-      var targetClassValues = dataset.GetDoubleValues(problemData.TargetVariable, rows);
-      var originalClassValues = model.GetEstimatedClassValues(dataset, rows);
+      var targetClassValues = dataset.GetDoubleValues(classificationProblemData.TargetVariable, rows);
+      var originalClassValues = classificationModel.GetEstimatedClassValues(dataset, rows);
       var qualityForImpactsCalculation = OnlineAccuracyCalculator.Calculate(targetClassValues, originalClassValues, out errorState);
       if (errorState != OnlineCalculatorError.None) qualityForImpactsCalculation = 0.0;
-
       return qualityForImpactsCalculation;
     }