Changeset 13221 for branches/DataAnalysis.ComplexityAnalyzer/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression

Timestamp:

11/17/15 15:57:52 (9 years ago)

Author:

mkommend

Message:

#2175: Removed unused features from complexity analysis.

Location:

branches/DataAnalysis.ComplexityAnalyzer/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4

Files:

• MultiObjective/PearsonRSquaredTreeComplexityEvaluator.cs (modified) (1 diff)
• MultiObjective/SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer.cs (modified) (3 diffs)
• SymbolicRegressionSolution.cs (modified) (4 diffs)

branches/DataAnalysis.ComplexityAnalyzer/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/MultiObjective/PearsonRSquaredTreeComplexityEvaluator.cs

@@ -65 +65 @@
       if (decimalPlaces >= 0)
         r2 = Math.Round(r2, decimalPlaces);
-      return new double[2] { r2, SymbolicDataAnalysisModelComplexityAnalyzer.CalculateComplexity(solution.Root.GetSubtree(0).GetSubtree(0)) };
+      return new double[2] { r2, SymbolicDataAnalysisModelComplexityCalculator.CalculateComplexity(solution) };
     }
 

branches/DataAnalysis.ComplexityAnalyzer/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/MultiObjective/SymbolicRegressionMultiObjectiveTrainingBestSolutionAnalyzer.cs

@@ -96 +96 @@
     public override IOperation Apply() {
       var operation = base.Apply();
-
       var paretoFront = TrainingBestSolutionsParameter.ActualValue;
 
@@ -131 +130 @@
       }
 
-
-
       qualityToTreeSize.Rows.Clear();
       var trainingRow = new ScatterPlotDataRow("Training NMSE", "", sizeParetoFront.Select(x => new Point2D<double>(x.Model.SymbolicExpressionTree.Length, x.TrainingNormalizedMeanSquaredError)));
-      trainingRow.VisualProperties.PointSize = 5;
+      trainingRow.VisualProperties.PointSize = 8;
       qualityToTreeSize.Rows.Add(trainingRow);
 
@@ -147 +144 @@
           sizeParetoFront.Select(x => new Point2D<double>(x.Model.SymbolicExpressionTree.Length,
           OnlineNormalizedMeanSquaredErrorCalculator.Calculate(targetValues, x.GetEstimatedValues(validationIndizes), out error))));
-        validationRow.VisualProperties.PointSize = 5;
+        validationRow.VisualProperties.PointSize = 7;
         qualityToTreeSize.Rows.Add(validationRow);
       }
-
-      double trainingArea = sizeParetoFront.Select(s => s.Model.SymbolicExpressionTree.Length * s.TrainingNormalizedMeanSquaredError).Average();
-      double testArea = sizeParetoFront.Select(s => s.Model.SymbolicExpressionTree.Length * s.TestNormalizedMeanSquaredError).Average();
-
-      ResultCollection paretoFrontResults;
-      if (!ResultCollection.TryGetValue("Pareto Front Results", out result)) {
-        paretoFrontResults = new ResultCollection();
-        ResultCollection.Add(new Result("Pareto Front Results", paretoFrontResults));
-      } else paretoFrontResults = (ResultCollection)result.Value;
-
-      DoubleValue trainingAreaResult, testAreaResult, areaDifferenceResult, avgTrainingNMSE, avgTestNMSE;
-      if (!paretoFrontResults.TryGetValue("Non Dominated Area (training)", out result)) {
-        trainingAreaResult = new DoubleValue();
-        paretoFrontResults.Add(new Result("Non Dominated Area (training)", trainingAreaResult));
-      } else trainingAreaResult = (DoubleValue)result.Value;
-      if (!paretoFrontResults.TryGetValue("Non Dominated Area (test)", out result)) {
-        testAreaResult = new DoubleValue();
-        paretoFrontResults.Add(new Result("Non Dominated Area (test)", testAreaResult));
-      } else testAreaResult = (DoubleValue)result.Value;
-      if (!paretoFrontResults.TryGetValue("Non Dominated Area Difference", out result)) {
-        areaDifferenceResult = new DoubleValue();
-        paretoFrontResults.Add(new Result("Non Dominated Area Difference", areaDifferenceResult));
-      } else areaDifferenceResult = (DoubleValue)result.Value;
-      if (!paretoFrontResults.TryGetValue("Average Training NMSE", out result)) {
-        avgTrainingNMSE = new DoubleValue();
-        paretoFrontResults.Add(new Result("Average Training NMSE", avgTrainingNMSE));
-      } else avgTrainingNMSE = (DoubleValue)result.Value;
-      if (!paretoFrontResults.TryGetValue("Average Test NMSE", out result)) {
-        avgTestNMSE = new DoubleValue();
-        paretoFrontResults.Add(new Result("Average Test NMSE", avgTestNMSE));
-      } else avgTestNMSE = (DoubleValue)result.Value;
-
-      trainingAreaResult.Value = trainingArea;
-      testAreaResult.Value = testArea;
-      areaDifferenceResult.Value = trainingArea - testArea;
-      avgTrainingNMSE.Value = sizeParetoFront.Select(s => s.TrainingNormalizedMeanSquaredError).Average();
-      avgTestNMSE.Value = sizeParetoFront.Select(s => s.TestNormalizedMeanSquaredError).Average();
 
       return operation;

branches/DataAnalysis.ComplexityAnalyzer/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SymbolicRegressionSolution.cs

@@ -64 +64 @@
     }
 
-    public double ModelComplexity {
-      get { return ((DoubleValue)this["Complexity"].Value).Value; }
-      private set { ((DoubleValue)this["Complexity"].Value).Value = value; }
-    }
-
     private ResultCollection EstimationLimitsResultCollection {
       get { return (ResultCollection)this[EstimationLimitsResultsResultName].Value; }
@@ -123 +118 @@
       estimationLimitResults.Add(new Result(TestNaNEvaluationsResultName, "", new IntValue()));
       Add(new Result(EstimationLimitsResultsResultName, "Results concerning the estimation limits of symbolic regression solution", estimationLimitResults));
-      Add(new Result("Complexity", "", new DoubleValue()));
       RecalculateResults();
     }
@@ -133 +127 @@
     [StorableHook(HookType.AfterDeserialization)]
     private void AfterDeserialization() {
-      if (!ContainsKey("Complexity")) Add(new Result("Complexity", "", new DoubleValue()));
       if (!ContainsKey(EstimationLimitsResultsResultName)) {
         ResultCollection estimationLimitResults = new ResultCollection();
@@ -156 +149 @@
       ModelLength = Model.SymbolicExpressionTree.Length;
       ModelDepth = Model.SymbolicExpressionTree.Depth;
-      ModelComplexity = SymbolicDataAnalysisModelComplexityAnalyzer.CalculateComplexity(Model.SymbolicExpressionTree.Root.GetSubtree(0).GetSubtree(0));
 
       EstimationLimits.Lower = Model.LowerEstimationLimit;