R help - Jan 2012 - rpart vs. tree and deviance calculations

Hi Everyone,

 

I'm working on building some classification trees, and up to this point
I've been using rpart.  However, I recently discovered the tree package,
and found that it had some useful functions (in particular deviance(),
which I would really like to use for my project).  I can't seem to find
an equivalent function for rpart.  I've considered using tree() in place
of rpart(), but I read an old post on this list that recommended using
rpart() as tree() was built for bug-checking in S.  So, to summarize:

 

Is there a way to compute deviance for an rpart object?

If not, would it be a bad idea to use tree() in place of rpart()?  My
rpart models are currently quite large and take ~ 15 minutes to run...

 

Thanks everyone!


	[[alternative HTML version deleted]]