Changeset 18104 for branches/3136_Structural_GP/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/StructuredSymbolicRegressionSingleObjectiveProblem.cs

Timestamp:

12/03/21 13:46:20 (2 years ago)

Author:

dpiringe

Message:

#3136

• overrode the method GetActualValue in ValueLookupParameter to get the default value when the execution context is null
• reverted the linear scaling logic for NMSESingleObjectiveConstraintsEvaluator
• in SymbolicRegressionConstantOptimizationEvaluator: removed the usage of GenerateRowsToEvaluate because it uses lookup parameters
• set the value of RelativeNumberOfEvaluatedSamplesParameter for SymbolicRegressionConstantOptimizationEvaluator in StructuredSymbolicRegressionSingleObjectiveProblem if Maximization = true and the SymbolicRegressionConstantOptimizationEvaluator is configured as evaluator
• added the SubFunctionSymbol in TreeToAutoDiffTermConverter

File:

• branches/3136_Structural_GP/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/StructuredSymbolicRegressionSingleObjectiveProblem.cs (modified) (1 diff)

branches/3136_Structural_GP/HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/StructuredSymbolicRegressionSingleObjectiveProblem.cs

@@ -194 +194 @@
       individual[SymbolicExpressionTreeName] = tree;
 
-      if(TreeEvaluatorParameter.Value is SymbolicRegressionConstantOptimizationEvaluator)
-        TreeEvaluatorParameter.Value.RandomParameter.Value = random;
+      // dpiringe: needed when Maximization = true
+      if (TreeEvaluatorParameter.Value is SymbolicRegressionConstantOptimizationEvaluator constantOptEvaluator) {
+        constantOptEvaluator.RandomParameter.Value = random;
+        constantOptEvaluator.RelativeNumberOfEvaluatedSamplesParameter.Value = 
+          (PercentValue)constantOptEvaluator.ConstantOptimizationRowsPercentage.Clone();
+      }
 
       return TreeEvaluatorParameter.Value.Evaluate(