Hi Everyone, Is there a way to get predict.rpart() to return the nodes reached by the new examples in addition to the predicted probabilities it already returns? In other words, I would like to know the leaf node in the tree object that each new example data drops down to. Thanks in advance for your help. Osei
Hi Everyone, Is there a way to get predict.rpart() to return the nodes reached by the new examples in addition to the predicted probabilities it already returns? In other words, I would like to know the leaf node in the tree object that each new example data drops down to. Thanks in advance for your help. Osei
Make your own copy of predict.rpart: mypredict <- rpart:::predict.rpart (Even better to grab the source code version, then you can start with a copy that has comments in it!) Edit it to add type='where': about 1/3 of the way down you will see that the routine has computed a variable "where"; that is what you want to return. Terry T. ---------- begin included message ---------- Is there a way to get predict.rpart() to return the nodes reached by the new examples in addition to the predicted probabilities it already returns? In other words, I would like to know the leaf node in the tree object that each new example data drops down to. Thanks in advance for your help.