[R-pkgs] `partykit': A Toolkit for Recursive Partytioning

Torsten Hothorn Torsten.Hothorn at R-project.org
Mon Oct 3 18:00:29 CEST 2011


New package `partykit': A Toolkit for Recursive Partytioning

The purpose of the package is to provide a toolkit with infrastructure for
representing, summarizing, and visualizing tree-structured regression and
classification models. Thus, the focus is not on _inferring_ such a
tree structure from data but to _represent_ a given tree so that
printing/plotting and computing predictions can be performed in a
standardized  way. In particular, this unified infrastructure can be
used for reading/coercing tree models from different sources
(packages `rpart', `RWeka', `PMML') yielding objects that share
functionality for `print()', `plot()', and `predict()' methods.

The impatient users will hopefully have fun with

install.packages("partykit")
library("partykit")
library("rpart")
### from ?rpart
fit <- rpart(Kyphosis ~ Age + Number + Start, data = kyphosis)
plot(as.party(fit))


Best,

Torsten & Achim



More information about the R-packages mailing list