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
|
\name{MDSplot}
\alias{MDSplot}
\title{Multi-dimensional Scaling Plot of Proximity matrix from randomForest}
\description{
Plot the scaling coordinates of the proximity matrix from randomForest.
}
\usage{
MDSplot(rf, fac, k=2, palette=NULL, pch=20, ...)
}
\arguments{
\item{rf}{an object of class \code{\link{randomForest}} that contains
the \code{proximity} component.}
\item{fac}{a factor that was used as response to train \code{rf}.}
\item{k}{number of dimensions for the scaling coordinates.}
\item{palette}{colors to use to distinguish the classes; length must
be the equal to the number of levels.}
\item{pch}{plotting symbols to use.}
\item{...}{other graphical parameters.}
}
\value{
The output of \code{\link[stats]{cmdscale}} on 1 - \code{rf$proximity} is
returned invisibly.
}
\note{
If \code{k > 2}, \code{\link{pairs}} is used to produce the
scatterplot matrix of the coordinates.
}
\seealso{\code{\link{randomForest}}}
\examples{
set.seed(1)
data(iris)
iris.rf <- randomForest(Species ~ ., iris, proximity=TRUE,
keep.forest=FALSE)
MDSplot(iris.rf, iris$Species)
## Using different symbols for the classes:
MDSplot(iris.rf, iris$Species, palette=rep(1, 3), pch=as.numeric(iris$Species))
}
\author{Robert Gentleman, with slight modifications by Andy Liaw}
\keyword{classif}
\keyword{tree}
|
