Changeset 15481

Timestamp:

11/21/17 15:40:19 (7 years ago)

Author:

gkronber

Message:

#2852: reverted changeset r15480 partially to fix unit test

Location:

trunk/sources

Files:

• HeuristicLab.Problems.DataAnalysis.Symbolic.Regression/3.4/SingleObjective/Evaluators/SymbolicRegressionConstantOptimizationEvaluator.cs (modified) (2 diffs)
• HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Converters/TreeToAutoDiffTermConverter.cs (modified) (1 diff)

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

@@ -213 +213 @@
       if (applyLinearScaling) {
         c = new double[initialConstants.Length + 2];
-        {
-          Array.Copy(initialConstants, 0, c, 0, initialConstants.Length);
-          c[c.Length - 2] = 0.0;
-          c[c.Length - 1] = 1.0;
-        }
+        c[0] = 0.0;
+        c[1] = 1.0;
+        Array.Copy(initialConstants, 0, c, 2, initialConstants.Length);
       } else {
         c = (double[])initialConstants.Clone();
@@ -273 +271 @@
       //retVal == -7  => constant optimization failed due to wrong gradient
       if (retVal != -7) {
-        if (applyLinearScaling) UpdateConstants(tree, c, updateVariableWeights);
-        else UpdateConstants(tree, c, updateVariableWeights);
+        if (applyLinearScaling) {
+          var tmp = new double[c.Length - 2];
+          Array.Copy(c, 2, tmp, 0, tmp.Length);
+          UpdateConstants(tree, tmp, updateVariableWeights);
+        } else UpdateConstants(tree, c, updateVariableWeights);
       }
       var quality = SymbolicRegressionSingleObjectivePearsonRSquaredEvaluator.Calculate(interpreter, tree, lowerEstimationLimit, upperEstimationLimit, problemData, rows, applyLinearScaling);

trunk/sources/HeuristicLab.Problems.DataAnalysis.Symbolic/3.4/Converters/TreeToAutoDiffTermConverter.cs

@@ -251 +251 @@
       if (node.Symbol is StartSymbol) {
         if (addLinearScalingTerms) {
-          var t = ConvertToAutoDiff(node.GetSubtree(0));
-          // scaling variables α, β are given at the end of the parameter vector
+          // scaling variables α, β are given at the beginning of the parameter vector
           var alpha = new AutoDiff.Variable();
           var beta = new AutoDiff.Variable();
           variables.Add(beta);
           variables.Add(alpha);
+          var t = ConvertToAutoDiff(node.GetSubtree(0));
           return t * alpha + beta;
         } else return ConvertToAutoDiff(node.GetSubtree(0));