[R] Surrogate splits for decision trees

[Prof Brian Ripley]

On Tue, 22 May 2001, Thomas Lumley wrote:

> On Tue, 22 May 2001, Clayton Springer wrote:
> 
> > Dear R,
> >
> > Short verse of the question:
> >
> > Is there R code which will calculate surrogate splits
> > and/or delta impurity for decision trees at each node?
> 
> package rpart
> 
> > Long Version:
> >
> > I have local, legacy code which I use to calculate my decision trees.
> > I would like to switch to R, but as I understand it surrogate splits
> > are not implemented.
> >
> > Surrogate splits and feature ranking are described in Breiman et al
> > "CART" 1984.  Curiously, they don't seem to appear anywhere else
> > in the literature.
> >
> 
> The rpart package contains tree code that does implement surrogate splits.
> It is not particularly extensible now, but the next version is planned to
> be very extensible.

And I am porting a pre-release of that version to R right now.


-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._