Changeset 17999 for trunk/HeuristicLab.Problems.DataAnalysis/3.4/DatasetExtensions.cs

Timestamp:

06/28/21 10:35:19 (11 months ago)

Author:

mkommend

Message:

#3129: Work on variable range calculation of datasets.

Renamed method to calculate variable ranges to a more appropriate method name.
Added possibility to ignore NaNs for range calculation.

File:

• trunk/HeuristicLab.Problems.DataAnalysis/3.4/DatasetExtensions.cs (modified) (1 diff)

trunk/HeuristicLab.Problems.DataAnalysis/3.4/DatasetExtensions.cs

@@ -96 +96 @@
     }
 
-    public static IntervalCollection GetIntervals(this IDataset dataset) {
-      IntervalCollection intervalCollection = new IntervalCollection();
-      foreach (var variable in dataset.DoubleVariables) { // intervals are only possible for double variables
-        var variableInterval = Interval.GetInterval(dataset.GetDoubleValues(variable));
-        intervalCollection.AddInterval(variable, variableInterval);
+    public static IntervalCollection GetVariableRanges(this IDataset dataset, bool ignoreNaNs = true) {
+      IntervalCollection variableRanges = new IntervalCollection();
+      foreach (var variable in dataset.DoubleVariables) { // ranges can only be calculated for double variables
+        var values = dataset.GetDoubleValues(variable);
+
+        if (ignoreNaNs) {
+          values = values.Where(v => !double.IsNaN(v));
+
+          if (!values.Any()) { //handle values with only NaNs explicitly
+            var emptyInterval = new Interval(double.NaN, double.NaN);
+            variableRanges.AddInterval(variable, emptyInterval);
+            continue;
+          }
+        }
+
+        var interval = Interval.GetInterval(values);
+        variableRanges.AddInterval(variable, interval);
       }
 
-      return intervalCollection;
+      return variableRanges;
     }