I trained rpart classification model using caret.
According to variable importance function, there is variable that is the second most important, but it doesn't appear in any of the splitting rules.
My question is what could be a possible explanation for this phenomena.
What special about this variable is that it is linearly correlated with the outcome.