Classification threshold view does not work correctly when there are no observations of a class in training or test partition
|Reported by:||abeham||Owned by:||mkommend|
|Priority:||low||Milestone:||HeuristicLab 3.3.x Backlog|
Description (last modified by abeham)
If a class has only datapoints in training, but not in test (presumably also vice-versa), the class is fully jittered in the classification threshold view.