Changeset 5275 for branches/DataAnalysis/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Analyzers/RegressionSolutionAnalyzer.cs

Timestamp:

01/11/11 15:03:46 (13 years ago)

Author:

gkronber

Message:

Merged changes from trunk to data analysis exploration branch and added fractional distance metric evaluator. #1142

File:

• branches/DataAnalysis/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Analyzers/RegressionSolutionAnalyzer.cs (modified) (9 diffs)

branches/DataAnalysis/HeuristicLab.Problems.DataAnalysis.Regression/3.3/Analyzers/RegressionSolutionAnalyzer.cs

@@ -21 +21 @@
 
 using System.Collections.Generic;
+using HeuristicLab.Common;
 using HeuristicLab.Core;
 using HeuristicLab.Data;
@@ -28 +29 @@
 using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
 using HeuristicLab.Problems.DataAnalysis.Evaluators;
-using HeuristicLab.Analysis;
 
 namespace HeuristicLab.Problems.DataAnalysis.Regression.Symbolic.Analyzers {
@@ -90 +90 @@
     #endregion
 
+
+    [StorableConstructor]
+    protected RegressionSolutionAnalyzer(bool deserializing) : base(deserializing) { }
+    protected RegressionSolutionAnalyzer(RegressionSolutionAnalyzer original, Cloner cloner) : base(original, cloner) { }
     public RegressionSolutionAnalyzer()
       : base() {
@@ -102 +106 @@
 
     [StorableHook(HookType.AfterDeserialization)]
-    private void Initialize() {
+    private void AfterDeserialization() {
       // backwards compatibility
       if (!Parameters.ContainsKey(GenerationsParameterName)) {
@@ -119 +123 @@
     }
 
-    public static void UpdateBestSolutionResults(DataAnalysisSolution bestSolution, DataAnalysisProblemData problemData, ResultCollection results, IntValue CurrentGeneration) {
-      var solution = bestSolution;
+    public static void UpdateBestSolutionResults(DataAnalysisSolution solution, DataAnalysisProblemData problemData, ResultCollection results, IntValue generation) {
       #region update R2,MSE, Rel Error
-      IEnumerable<double> trainingValues = problemData.Dataset.GetEnumeratedVariableValues(
-        problemData.TargetVariable.Value,
-        problemData.TrainingSamplesStart.Value,
-        problemData.TrainingSamplesEnd.Value);
-      IEnumerable<double> testValues = problemData.Dataset.GetEnumeratedVariableValues(
-        problemData.TargetVariable.Value,
-        problemData.TestSamplesStart.Value,
-        problemData.TestSamplesEnd.Value);
+      IEnumerable<double> trainingValues = problemData.Dataset.GetEnumeratedVariableValues(problemData.TargetVariable.Value, problemData.TrainingIndizes);
+      IEnumerable<double> testValues = problemData.Dataset.GetEnumeratedVariableValues(problemData.TargetVariable.Value, problemData.TestIndizes);
       OnlineMeanSquaredErrorEvaluator mseEvaluator = new OnlineMeanSquaredErrorEvaluator();
       OnlineMeanAbsolutePercentageErrorEvaluator relErrorEvaluator = new OnlineMeanAbsolutePercentageErrorEvaluator();
       OnlinePearsonsRSquaredEvaluator r2Evaluator = new OnlinePearsonsRSquaredEvaluator();
+
       #region training
       var originalEnumerator = trainingValues.GetEnumerator();
@@ -145 +143 @@
       double trainingRelError = relErrorEvaluator.MeanAbsolutePercentageError;
       #endregion
+
       mseEvaluator.Reset();
       relErrorEvaluator.Reset();
       r2Evaluator.Reset();
+
       #region test
       originalEnumerator = testValues.GetEnumerator();
@@ -160 +160 @@
       double testRelError = relErrorEvaluator.MeanAbsolutePercentageError;
       #endregion
+
       if (results.ContainsKey(BestSolutionResultName)) {
         results[BestSolutionResultName].Value = solution;
@@ -168 +169 @@
         results[BestSolutionTrainingRelativeError].Value = new DoubleValue(trainingRelError);
         results[BestSolutionTestRelativeError].Value = new DoubleValue(testRelError);
-        if (CurrentGeneration != null) // this check is needed because linear regression solutions do not have a generations parameter
-          results[BestSolutionGeneration].Value = new IntValue(CurrentGeneration.Value);
-        var solutionQualityTable = (DataTable)results["Best solution quality table"].Value;
-        solutionQualityTable.Rows["Training R²"].Values.Add(trainingR2);
-        solutionQualityTable.Rows["Training MSE"].Values.Add(trainingMse);
-        solutionQualityTable.Rows["Test R²"].Values.Add(testR2);
-        solutionQualityTable.Rows["Test MSE"].Values.Add(testMse);
+        if (generation != null) // this check is needed because linear regression solutions do not have a generations parameter
+          results[BestSolutionGeneration].Value = new IntValue(generation.Value);
       } else {
         results.Add(new Result(BestSolutionResultName, solution));
@@ -183 +179 @@
         results.Add(new Result(BestSolutionTrainingRelativeError, new DoubleValue(trainingRelError)));
         results.Add(new Result(BestSolutionTestRelativeError, new DoubleValue(testRelError)));
-        if (CurrentGeneration != null)
-          results.Add(new Result(BestSolutionGeneration, new IntValue(CurrentGeneration.Value)));
-        var solutionQualityTable = new DataTable("Best solution quality table");
-        solutionQualityTable.Rows.Add(new DataRow("Training R²"));
-        solutionQualityTable.Rows.Add(new DataRow("Training MSE"));
-        solutionQualityTable.Rows.Add(new DataRow("Test R²"));
-        solutionQualityTable.Rows.Add(new DataRow("Test MSE"));
-        solutionQualityTable.Rows["Training R²"].Values.Add(trainingR2);
-        solutionQualityTable.Rows["Training MSE"].Values.Add(trainingMse);
-        solutionQualityTable.Rows["Test R²"].Values.Add(testR2);
-        solutionQualityTable.Rows["Test MSE"].Values.Add(testMse);
-        results.Add(new Result("Best solution quality table", solutionQualityTable));
+        if (generation != null)
+          results.Add(new Result(BestSolutionGeneration, new IntValue(generation.Value)));
       }
       #endregion