Changeset 16034 for branches/2904_CalculateImpacts/3.4/Implementation/Regression/RegressionSolutionVariableImpactsCalculator.cs

Timestamp:

07/30/18 16:59:47 (6 years ago)

Author:

fholzing

Message:

#2904: Removed callback, adapted both view and calculator.

File:

• branches/2904_CalculateImpacts/3.4/Implementation/Regression/RegressionSolutionVariableImpactsCalculator.cs (modified) (6 diffs)

branches/2904_CalculateImpacts/3.4/Implementation/Regression/RegressionSolutionVariableImpactsCalculator.cs

@@ -106 +106 @@
       ReplacementMethodEnum replacementMethod = ReplacementMethodEnum.Shuffle,
       FactorReplacementMethodEnum factorReplacementMethod = FactorReplacementMethodEnum.Best,
-      DataPartitionEnum data = DataPartitionEnum.Training,
-      Func<double, string, bool> progressCallback = null) {
-      return CalculateImpacts(solution.Model, solution.ProblemData, solution.EstimatedValues, replacementMethod, factorReplacementMethod, data, progressCallback);
+      DataPartitionEnum data = DataPartitionEnum.Training) {
+      return CalculateImpacts(solution.Model, solution.ProblemData, solution.EstimatedValues, replacementMethod, factorReplacementMethod, data);
     }
 
@@ -117 +116 @@
       ReplacementMethodEnum replacementMethod = ReplacementMethodEnum.Shuffle,
       FactorReplacementMethodEnum factorReplacementMethod = FactorReplacementMethodEnum.Best,
-      DataPartitionEnum data = DataPartitionEnum.Training,
-      Func<double, string, bool> progressCallback = null) {
+      DataPartitionEnum data = DataPartitionEnum.Training) {
       IEnumerable<int> rows;
 
@@ -135 +133 @@
       }
 
-      return CalculateImpacts(model, problemData, estimatedValues, rows, replacementMethod, factorReplacementMethod, progressCallback);
+      return CalculateImpacts(model, problemData, estimatedValues, rows, replacementMethod, factorReplacementMethod);
+    }
+
+    public static double CalculateImpact(string variableName, IRegressionModel model, IRegressionProblemData problemData, IEnumerable<double> estimatedValues, DataPartitionEnum dataPartition, ReplacementMethodEnum replMethod, FactorReplacementMethodEnum factorReplMethod) {
+      double impact = 0;
+
+      IEnumerable<int> rows;
+      switch (dataPartition) {
+        case DataPartitionEnum.All:
+          rows = problemData.AllIndices;
+          break;
+        case DataPartitionEnum.Test:
+          rows = problemData.TestIndices;
+          break;
+        case DataPartitionEnum.Training:
+          rows = problemData.TrainingIndices;
+          break;
+        default:
+          throw new NotSupportedException("DataPartition not supported");
+      }
+
+      OnlineCalculatorError error;
+      IEnumerable<double> targetValuesPartition = rows.Select(v => problemData.TargetVariableValues.ElementAt(v));
+      IEnumerable<double> estimatedValuesPartition = rows.Select(v => estimatedValues.ElementAt(v));
+      var originalCalculatorValue = CalculateVariableImpact(targetValuesPartition, estimatedValuesPartition, out error);
+      if (error != OnlineCalculatorError.None) throw new InvalidOperationException("Error during calculation.");
+
+
+      var modifiableDataset = ((Dataset)(problemData.Dataset).Clone()).ToModifiable();
+
+      // calculate impacts for double variables
+      if (problemData.Dataset.VariableHasType<double>(variableName)) {
+        impact = CalculateImpactForNumericalVariables(variableName, model, modifiableDataset, rows, targetValuesPartition, originalCalculatorValue, replMethod);
+      } else if (problemData.Dataset.VariableHasType<string>(variableName)) {
+        impact = CalculateImpactForFactorVariables(variableName, model, problemData.Dataset, modifiableDataset, rows, targetValuesPartition, originalCalculatorValue, factorReplMethod);
+      } else {
+        throw new NotSupportedException("Variable not supported");
+      }
+      return impact;
     }
 
@@ -144 +180 @@
      IEnumerable<int> rows,
      ReplacementMethodEnum replacementMethod = ReplacementMethodEnum.Shuffle,
-     FactorReplacementMethodEnum factorReplacementMethod = FactorReplacementMethodEnum.Best,
-     Func<double, string, bool> progressCallback = null) {
+     FactorReplacementMethodEnum factorReplacementMethod = FactorReplacementMethodEnum.Best) {
       //Calculate original quality-values (via calculator, default is R²)
       OnlineCalculatorError error;
@@ -157 +192 @@
       var allowedInputVariables = problemData.Dataset.VariableNames.Where(v => inputvariables.Contains(v)).ToList();
 
-      int curIdx = 0;
-      int count = allowedInputVariables.Count(v => problemData.Dataset.VariableHasType<double>(v) || problemData.Dataset.VariableHasType<string>(v));
-
       foreach (var inputVariable in allowedInputVariables) {
-        //Report the current progress in percent. If the callback returns true, it means the execution shall be stopped
-        if (progressCallback != null) {
-          curIdx++;
-          if (progressCallback((double)curIdx / count, string.Format("Calculating impact for variable {0} ({1} of {2})", inputVariable, curIdx, count))) { return null; }
-        }
         impacts[inputVariable] = CalculateImpact(inputVariable, model, problemData.Dataset, rows, targetValuesPartition, originalCalculatorValue, replacementMethod, factorReplacementMethod);
       }
@@ -204 +231 @@
       }
       return impact;
-    }
-
-    private static void PrepareData(IEnumerable<int> rows,
-      IRegressionProblemData problemData,
-      IEnumerable<double> estimatedValues,
-      out IEnumerable<double> targetValues,
-      out double originalValue) {
-      OnlineCalculatorError error;
-
-      var targetVariableValueList = problemData.TargetVariableValues.ToList();
-      targetValues = rows.Select(v => targetVariableValueList.ElementAt(v));
-      var estimatedValuesPartition = rows.Select(v => estimatedValues.ElementAt(v));
-      originalValue = CalculateVariableImpact(targetValues, estimatedValuesPartition, out error);
-
-      if (error != OnlineCalculatorError.None) throw new InvalidOperationException("Error during calculation.");
     }