Changeset 7677 for trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/Evaluators/SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator.cs

Timestamp:

03/30/12 22:48:32 (13 years ago)

Author:

mkommend

Message:

#1788: Implemente new symbolic regression evaluators.

File:

• trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/Evaluators/SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator.cs (copied) (copied from trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/Evaluators/SymbolicRegressionSingleObjectiveMeanSquaredErrorEvaluator.cs) (4 diffs)

trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/Evaluators/SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator.cs

@@ -22 +22 @@
 using System;
 using System.Collections.Generic;
-using System.Linq;
 using HeuristicLab.Common;
 using HeuristicLab.Core;
@@ -30 +29 @@
 
 namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Regression {
-  [Item("Mean squared error Evaluator", "Calculates the mean squared error of a symbolic regression solution.")]
+  [Item("Mean absolute error Evaluator", "Calculates the mean absolute error of a symbolic regression solution.")]
   [StorableClass]
-  public class SymbolicRegressionSingleObjectiveMeanSquaredErrorEvaluator : SymbolicRegressionSingleObjectiveEvaluator {
+  public class SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator : SymbolicRegressionSingleObjectiveEvaluator {
     public override bool Maximization { get { return false; } }
-    [ThreadStatic]
-    private static double[] estimatedValuesCache;
-
     [StorableConstructor]
-    protected SymbolicRegressionSingleObjectiveMeanSquaredErrorEvaluator(bool deserializing) : base(deserializing) { }
-    protected SymbolicRegressionSingleObjectiveMeanSquaredErrorEvaluator(SymbolicRegressionSingleObjectiveMeanSquaredErrorEvaluator original, Cloner cloner)
+    protected SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator(bool deserializing) : base(deserializing) { }
+    protected SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator(SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator original, Cloner cloner)
       : base(original, cloner) {
     }
     public override IDeepCloneable Clone(Cloner cloner) {
-      return new SymbolicRegressionSingleObjectiveMeanSquaredErrorEvaluator(this, cloner);
+      return new SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator(this, cloner);
     }
-    public SymbolicRegressionSingleObjectiveMeanSquaredErrorEvaluator() : base() { }
+    public SymbolicRegressionSingleObjectiveMeanAbsoluteErrorEvaluator() : base() { }
 
     public override IOperation Apply() {
@@ -59 +55 @@
     public static double Calculate(ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, ISymbolicExpressionTree solution, double lowerEstimationLimit, double upperEstimationLimit, IRegressionProblemData problemData, IEnumerable<int> rows, bool applyLinearScaling) {
       IEnumerable<double> estimatedValues = interpreter.GetSymbolicExpressionTreeValues(solution, problemData.Dataset, rows);
-      IEnumerable<double> originalValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, rows);
+      IEnumerable<double> targetValues = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, rows);
       IEnumerable<double> boundedEstimatedValues = estimatedValues.LimitToRange(lowerEstimationLimit, upperEstimationLimit);
       OnlineCalculatorError errorState;
@@ -65 +61 @@
       double mse;
       if (applyLinearScaling) {
-        //int i = 0;
-        //int rowCount = rows.Count();
-        //if (estimatedValuesCache == null) estimatedValuesCache = new double[problemData.Dataset.Rows];
-        //foreach (var value in boundedEstimatedValues) {
-        //  estimatedValuesCache[i] = value;
-        //  i++;
-        //}
-
-        double alpha, beta;
-        OnlineLinearScalingParameterCalculator.Calculate(boundedEstimatedValues, originalValues, out alpha, out beta, out errorState);
-        mse = OnlineMeanSquaredErrorCalculator.Calculate(originalValues, boundedEstimatedValues.Select(value => value * beta + alpha), out errorState);
+        var maeCalculator = new OnlineMeanAbsoluteErrorCalculator();
+        CalculateWithScaling(targetValues, boundedEstimatedValues, maeCalculator, problemData.Dataset.Rows);
+        errorState = maeCalculator.ErrorState;
+        mse = maeCalculator.MeanAbsoluteError;
       } else
-        mse = OnlineMeanSquaredErrorCalculator.Calculate(originalValues, boundedEstimatedValues, out errorState);
+        mse = OnlineMeanSquaredErrorCalculator.Calculate(targetValues, boundedEstimatedValues, out errorState);
 
       if (errorState != OnlineCalculatorError.None) return Double.NaN;