1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
|
\name{rpart}
\alias{rpart}
\alias{rpartcallback}
\title{
Recursive Partitioning and Regression Trees
}
\description{
Fit a \code{rpart} model
}
\usage{
rpart(formula, data, weights, subset, na.action = na.rpart, method,
model = FALSE, x = FALSE, y = TRUE, parms, control, cost, \dots)
}
\arguments{
\item{formula}{
a formula, as in the \code{lm} function.
}
\item{data}{
an optional data frame in which to interpret the variables named in the
formula
}
\item{weights}{
optional case weights.
}
\item{subset}{
optional expression saying that only a subset of the rows of the data
should be used in the fit.
}
\item{na.action}{
The default action deletes all observations for which \code{y} is missing,
but keeps those in which one or more predictors are missing.
}
\item{method}{
one of \code{"anova"}, \code{"poisson"}, \code{"class"} or \code{"exp"}.
If \code{method} is missing then the routine tries to make an intelligent guess.
If \code{y} is a survival object, then \code{method="exp"} is assumed,
if \code{y} has 2 columns then \code{method="poisson"} is assumed,
if \code{y} is a factor then \code{method="class"} is assumed, otherwise \code{method="anova"}
is assumed. It is wisest to specify the method directly, especially as
more criteria are added to the function.
Alternatively, \code{method} can be a list of functions named
\code{init}, \code{split} and \code{eval}. Examples are given in
the file \file{tests/usersplits.R} in the sources.
}
\item{model}{
if logical: keep a copy of the model frame in the result? If the input
value for \code{model} is a model frame (likely from an earlier call to
the \code{rpart} function), then this frame is used rather than
constructing new data.
}
\item{x}{
keep a copy of the \code{x} matrix in the result.
}
\item{y}{
keep a copy of the dependent variable in the result. If missing and
\code{model} is supplied this defaults to \code{FALSE}.
}
\item{parms}{
optional parameters for the splitting function.
Anova splitting has no parameters.
Poisson splitting has a single parameter, the coefficient of variation of
the prior distribution on the rates. The default value is 1.
Exponential splitting has the same parameter as Poisson.
For classification splitting, the list can contain any of:
the vector of prior probabilities (component \code{prior}), the loss matrix
(component \code{loss}) or the splitting index (component \code{split}). The
priors must be positive and sum to 1. The loss matrix must have zeros
on the diagonal and positive off-diagonal elements. The splitting
index can be \code{gini} or \code{information}. The default priors are
proportional to the data counts, the losses default to 1,
and the split defaults to \code{gini}.
}
\item{control}{
options that control details of the \code{rpart} algorithm.
}
\item{cost}{
a vector of non-negative costs, one for each variable in the
model. Defaults to one for all variables. These are scalings to be
applied when considering splits, so the improvement on splitting on a
variable is divided by its cost in deciding which split to choose.
}
\item{\dots}{
arguments to \code{rpart.control} may also be specified in the call to
\code{rpart}. They are checked against the list of valid arguments.
}
}
\value{
an object of class \code{rpart}, a superset of class \code{tree}.
}
\details{
This differs from the \code{tree} function mainly in its handling of surrogate
variables. In most details it follows Breiman et. al. quite closely.
}
\references{
Breiman, Friedman, Olshen, and Stone. (1984)
\emph{Classification and Regression Trees.}
Wadsworth.
}
\seealso{
\code{\link{rpart.control}}, \code{\link{rpart.object}},
\code{\link{summary.rpart}}, \code{\link{print.rpart}}
}
\examples{
fit <- rpart(Kyphosis ~ Age + Number + Start, data=kyphosis)
fit2 <- rpart(Kyphosis ~ Age + Number + Start, data=kyphosis,
parms=list(prior=c(.65,.35), split='information'))
fit3 <- rpart(Kyphosis ~ Age + Number + Start, data=kyphosis,
control=rpart.control(cp=.05))
par(mfrow=c(1,2), xpd=NA) # otherwise on some devices the text is clipped
plot(fit)
text(fit, use.n=TRUE)
plot(fit2)
text(fit2, use.n=TRUE)
}
\keyword{tree}
|
