R help - Sep 2005 - Finding a decision tree's leaf node from a new value

Dear mailinglist members,

I have the following problem: I run a decision tree using the rpart function
and, afterwords, I try to find to which leaf node a new register (not used to
build the decision tree) belongs to.

I will try to explain better:

rpart.tree <- rpart(target.value ~., data)
leaf.node <- new.function(rpart.tree, new.register)

The new register has all the explanatory values used to build the tree but has
not the target value.
The new.function is the function I am trying to write and shall return the
leaf.node from the decision tree where it should belong.

What I need ( and I am am not able to) is to reproduce the decision path over my
new.register.

Thanks in advance

Joao Moreira
	[[alternative HTML version deleted]]

It sounds to me like you're looking for the function predict.rpart().

Hope this helps,

Stephen

On Sep 9, 2005, at 8:45 AM, Jo??o Mendes Moreira wrote:
> Dear mailinglist members,
>
> I have the following problem: I run a decision tree using the rpart 
> function and, afterwords, I try to find to which leaf node a new 
> register (not used to build the decision tree) belongs to.
>
> I will try to explain better:
>
> rpart.tree <- rpart(target.value ~., data)
> leaf.node <- new.function(rpart.tree, new.register)
>
> The new register has all the explanatory values used to build the tree 
> but has not the target value.
> The new.function is the function I am trying to write and shall return 
> the leaf.node from the decision tree where it should belong.
>
> What I need ( and I am am not able to) is to reproduce the decision 
> path over my new.register.
>
> Thanks in advance
>
> Joao Moreira