Changeset 5331 for trunk/sources/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic

Timestamp:

01/19/11 15:07:45 (14 years ago)

Author:

mkommend

Message:

Added ApplyScalingParameter to RegressionAnalyzers (ticket #1369).

Location:

trunk/sources/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic

Files:

• Analyzers/FixedValidationBestScaledSymbolicRegressionSolutionAnalyzer.cs (modified) (8 diffs)
• Analyzers/TrainingBestScaledSymbolicRegressionSolutionAnalyzer.cs (modified) (8 diffs)
• SymbolicRegressionProblem.cs (modified) (1 diff)

trunk/sources/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic/Analyzers/FixedValidationBestScaledSymbolicRegressionSolutionAnalyzer.cs

@@ -20 +20 @@
 #endregion
 
-using System.Collections.Generic;
-using System.Linq;
 using HeuristicLab.Analysis;
 using HeuristicLab.Common;
@@ -27 +25 @@
 using HeuristicLab.Data;
 using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
-using HeuristicLab.Operators;
 using HeuristicLab.Optimization;
 using HeuristicLab.Parameters;
@@ -40 +37 @@
   [StorableClass]
   public sealed class FixedValidationBestScaledSymbolicRegressionSolutionAnalyzer : SymbolicRegressionValidationAnalyzer, ISymbolicRegressionAnalyzer {
+    private const string ApplyLinearScalingParameterName = "ApplyLinearScaling";
     private const string MaximizationParameterName = "Maximization";
     private const string CalculateSolutionComplexityParameterName = "CalculateSolutionComplexity";
@@ -84 +82 @@
       get { return (ILookupParameter<DataTable>)Parameters[VariableFrequenciesParameterName]; }
     }
-
+    public IValueLookupParameter<BoolValue> ApplyLinearScalingParameter {
+      get { return (IValueLookupParameter<BoolValue>)Parameters[ApplyLinearScalingParameterName]; }
+    }
     #endregion
     #region properties
@@ -114 +114 @@
       set { BestSolutionHeightParameter.ActualValue = value; }
     }
-
+    public BoolValue ApplyLinearScaling {
+      get { return ApplyLinearScalingParameter.ActualValue; }
+      set { ApplyLinearScalingParameter.ActualValue = value; }
+    }
     #endregion
 
@@ -122 +125 @@
     public FixedValidationBestScaledSymbolicRegressionSolutionAnalyzer()
       : base() {
+      Parameters.Add(new ValueLookupParameter<BoolValue>(ApplyLinearScalingParameterName, "The switch determines if the best solution should be linearly scaled on the whole training set.", new BoolValue(true)));
       Parameters.Add(new LookupParameter<BoolValue>(MaximizationParameterName, "The direction of optimization."));
-      Parameters.Add(new ValueParameter<BoolValue>(CalculateSolutionComplexityParameterName, "Determines if the length and height of the validation best solution should be calculated.", new BoolValue(false)));
+      Parameters.Add(new ValueParameter<BoolValue>(CalculateSolutionComplexityParameterName, "Determines if the length and height of the validation best solution should be calculated.", new BoolValue(true)));
       Parameters.Add(new LookupParameter<SymbolicRegressionSolution>(BestSolutionParameterName, "The best symbolic regression solution."));
       Parameters.Add(new LookupParameter<IntValue>(GenerationsParameterName, "The number of generations calculated so far."));
@@ -155 +159 @@
       if (!Parameters.ContainsKey(BestSolutionHeightParameterName)) {
         Parameters.Add(new LookupParameter<IntValue>(BestSolutionHeightParameterName, "The height of the best symbolic regression solution."));
+      }
+      if (!Parameters.ContainsKey(ApplyLinearScalingParameterName)) {
+        Parameters.Add(new ValueLookupParameter<BoolValue>(ApplyLinearScalingParameterName, "The switch determines if the best solution should be linearly scaled on the whole training set.", new BoolValue(true)));
       }
       #endregion
@@ -182 +189 @@
         string targetVariable = ProblemData.TargetVariable.Value;
 
-        // calculate scaling parameters and only for the best tree using the full training set
-        double alpha, beta;
-        SymbolicRegressionScaledMeanSquaredErrorEvaluator.Calculate(SymbolicExpressionTreeInterpreter, bestTree,
-          lowerEstimationLimit, upperEstimationLimit,
-          ProblemData.Dataset, targetVariable,
-          ProblemData.TrainingIndizes, out beta, out alpha);
-
-        // scale tree for solution
-        var scaledTree = SymbolicRegressionSolutionLinearScaler.Scale(bestTree, alpha, beta);
+        if (ApplyLinearScaling.Value) {
+          // calculate scaling parameters and only for the best tree using the full training set
+          double alpha, beta;
+          SymbolicRegressionScaledMeanSquaredErrorEvaluator.Calculate(SymbolicExpressionTreeInterpreter, bestTree,
+            lowerEstimationLimit, upperEstimationLimit,
+            ProblemData.Dataset, targetVariable,
+            ProblemData.TrainingIndizes, out beta, out alpha);
+
+          // scale tree for solution
+          bestTree = SymbolicRegressionSolutionLinearScaler.Scale(bestTree, alpha, beta);
+        }
         var model = new SymbolicRegressionModel((ISymbolicExpressionTreeInterpreter)SymbolicExpressionTreeInterpreter.Clone(),
-          scaledTree);
+          bestTree);
         var solution = new SymbolicRegressionSolution((DataAnalysisProblemData)ProblemData.Clone(), model, lowerEstimationLimit, upperEstimationLimit);
         solution.Name = BestSolutionParameterName;

trunk/sources/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic/Analyzers/TrainingBestScaledSymbolicRegressionSolutionAnalyzer.cs

@@ -22 +22 @@
 using System.Collections.Generic;
 using System.Linq;
-using HeuristicLab.Analysis;
 using HeuristicLab.Common;
 using HeuristicLab.Core;
@@ -31 +30 @@
 using HeuristicLab.Parameters;
 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
+using HeuristicLab.Problems.DataAnalysis.Evaluators;
 using HeuristicLab.Problems.DataAnalysis.Symbolic;
-using HeuristicLab.Problems.DataAnalysis.Evaluators;
 
 namespace HeuristicLab.Problems.DataAnalysis.Regression.Symbolic.Analyzers {
@@ -41 +40 @@
   [StorableClass]
   public sealed class TrainingBestScaledSymbolicRegressionSolutionAnalyzer : SingleSuccessorOperator, ISymbolicRegressionAnalyzer {
+    private const string ApplyLinearScalingParameterName = "ApplyLinearScaling";
     private const string SymbolicExpressionTreeParameterName = "SymbolicExpressionTree";
     private const string QualityParameterName = "Quality";
@@ -128 +128 @@
       get { return (ILookupParameter<ResultCollection>)Parameters[ResultsParameterName]; }
     }
+    public IValueLookupParameter<BoolValue> ApplyLinearScalingParameter {
+      get { return (IValueLookupParameter<BoolValue>)Parameters[ApplyLinearScalingParameterName]; }
+    }
     #endregion
     #region properties
@@ -205 +208 @@
       get { return BestSolutionTestRelativeErrorParameter.ActualValue; }
       set { BestSolutionTestRelativeErrorParameter.ActualValue = value; }
+    }
+    public BoolValue ApplyLinearScaling {
+      get { return ApplyLinearScalingParameter.ActualValue; }
+      set { ApplyLinearScalingParameter.ActualValue = value; }
     }
     #endregion
@@ -213 +220 @@
     public TrainingBestScaledSymbolicRegressionSolutionAnalyzer()
       : base() {
+      Parameters.Add(new ValueLookupParameter<BoolValue>(ApplyLinearScalingParameterName, "The switch determines if the best solution should be linearly scaled on the whole training set.", new BoolValue(true)));
       Parameters.Add(new LookupParameter<BoolValue>(MaximizationParameterName, "The direction of optimization."));
       Parameters.Add(new ScopeTreeLookupParameter<SymbolicExpressionTree>(SymbolicExpressionTreeParameterName, "The symbolic expression trees to analyze."));
       Parameters.Add(new ScopeTreeLookupParameter<DoubleValue>(QualityParameterName, "The qualities of the symbolic expression trees to analyze."));
-      Parameters.Add(new ValueParameter<BoolValue>(CalculateSolutionComplexityParameterName, "Determines if the length and height of the training best solution should be calculated.", new BoolValue(false)));
-      Parameters.Add(new ValueParameter<BoolValue>(CalculateSolutionAccuracyParameterName, "Determines if the accuracy of the training best solution on the training and test set should be calculated.", new BoolValue(false)));
+      Parameters.Add(new ValueParameter<BoolValue>(CalculateSolutionComplexityParameterName, "Determines if the length and height of the training best solution should be calculated.", new BoolValue(true)));
+      Parameters.Add(new ValueParameter<BoolValue>(CalculateSolutionAccuracyParameterName, "Determines if the accuracy of the training best solution on the training and test set should be calculated.", new BoolValue(true)));
       Parameters.Add(new ValueLookupParameter<ISymbolicExpressionTreeInterpreter>(SymbolicExpressionTreeInterpreterParameterName, "The interpreter that should be used for the analysis of symbolic expression trees."));
       Parameters.Add(new ValueLookupParameter<DataAnalysisProblemData>(ProblemDataParameterName, "The problem data for which the symbolic expression tree is a solution."));
@@ -241 +249 @@
 
     [StorableHook(HookType.AfterDeserialization)]
-    private void AfterDeserialization() { }
+    private void AfterDeserialization() {
+      if (!Parameters.ContainsKey(ApplyLinearScalingParameterName)) {
+        Parameters.Add(new ValueLookupParameter<BoolValue>(ApplyLinearScalingParameterName, "The switch determines if the best solution should be linearly scaled on the whole training set.", new BoolValue(true)));
+      }
+    }
 
     public override IOperation Apply() {
@@ -269 +281 @@
         string targetVariable = ProblemData.TargetVariable.Value;
 
-        // calculate scaling parameters and only for the best tree using the full training set
-        double alpha, beta;
-        SymbolicRegressionScaledMeanSquaredErrorEvaluator.Calculate(SymbolicExpressionTreeInterpreter, bestTree,
-          lowerEstimationLimit, upperEstimationLimit,
-          ProblemData.Dataset, targetVariable,
-          ProblemData.TrainingIndizes, out beta, out alpha);
-
-        // scale tree for solution
-        var scaledTree = SymbolicRegressionSolutionLinearScaler.Scale(bestTree, alpha, beta);
+        if (ApplyLinearScaling.Value) {
+          // calculate scaling parameters and only for the best tree using the full training set
+          double alpha, beta;
+          SymbolicRegressionScaledMeanSquaredErrorEvaluator.Calculate(SymbolicExpressionTreeInterpreter, bestTree,
+            lowerEstimationLimit, upperEstimationLimit,
+            ProblemData.Dataset, targetVariable,
+            ProblemData.TrainingIndizes, out beta, out alpha);
+
+          // scale tree for solution
+          bestTree = SymbolicRegressionSolutionLinearScaler.Scale(bestTree, alpha, beta);
+        }
         var model = new SymbolicRegressionModel((ISymbolicExpressionTreeInterpreter)SymbolicExpressionTreeInterpreter.Clone(),
-          scaledTree);
+          bestTree);
         var solution = new SymbolicRegressionSolution((DataAnalysisProblemData)ProblemData.Clone(), model, lowerEstimationLimit, upperEstimationLimit);
         solution.Name = BestSolutionParameterName;

trunk/sources/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Symbolic/SymbolicRegressionProblem.cs

@@ -148 +148 @@
       AddOperator(new FixedValidationBestScaledSymbolicRegressionSolutionAnalyzer());
       AddOperator(new SymbolicRegressionOverfittingAnalyzer());
+      AddOperator(new TrainingBestScaledSymbolicRegressionSolutionAnalyzer());
       ParameterizeAnalyzers();
     }