How to get the decision tree built at an iteration of the Ada boosting package in R?

Question

I have built a boosting model with 50 decision trees using the "ada" package in R. How can I print the decision tree built at a iteration of "ada"?

I can use the "print" function to display the detail of a "rpart" object, is there any similar function to print the decision trees built in an "ada" object?

Thank you very much!

My sample code is like below:

    library(ada)
    data("soldat")
    model <- ada(y ~ .
               , data=soldat
               , loss="e"
               , type="discrete"
               , iter=50
               , control=rpart.control(cp=-1, maxdepth=2))

Can you show a bit of code, a reproducible example perhaps? – Roman Luštrik Aug 12 '15 at 07:07 — Roman Luštrik, Aug 12 '15 at 07:07

score 2 · Accepted Answer · answered Aug 12 '15 at 08:43

2

The trees are stored in the list model$model$trees These are rpart objects so you can use print, plot etc., just as when calling rpart.

So, for instance to plot the 5th tree:

plot(model$model$trees[[5]])
text(model$model$trees[[5]])

answered Aug 12 '15 at 08:43

nico

50,859
17
87
112

How to get the decision tree built at an iteration of the Ada boosting package in R?

1 Answers1