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
|
\name{to.data.frame}
\alias{to.data.frame}
\title{Convert an Array into a Data Frame}
\description{
\code{to.data.frame} converts an array into a data frame, in such a way
that a chosen dimensional extent forms variables in the data frame.
The elements of the array must be either atomic, data frames
with matching variables, or coercable into such data frames.
}
\usage{
to.data.frame(X,as.vars=1,name="Freq")
}
\arguments{
\item{X}{an array.}
\item{as.vars}{a numeric value or a character string.
If it is a numeric value then it indicates the dimensional extend
which defines the variables. If it is a character string then it is
matched against the names of the dimenstional extents. This is
applicable e.g. if \code{X} is a contingency table and the
dimensional extents are named after the cross-classified factors.
Takes effect only if \code{X} is
an atomic array. If \code{as.vars} equals zero, a new variable
is created that contains the values of the array, that is,
\code{to.data.frame} acts on the array \code{X}
like \code{as.data.frame(as.table(X))}
}
\item{name}{a character string; the name of the variable
created if \code{X} is an atomic array and \code{as.vars} equals zero.
}
}
\value{
A data frame.
}
\examples{
berkeley <- Aggregate(Table(Admit,Freq)~.,data=UCBAdmissions)
berktest1 <- By(~Dept+Gender,
glm(cbind(Admitted,Rejected)~1,family="binomial"),
data=berkeley)
berktest2 <- By(~Dept,
glm(cbind(Admitted,Rejected)~Gender,family="binomial"),
data=berkeley)
Stest1 <- Lapply(berktest2,function(x)predict(x,,se.fit=TRUE)[c("fit","se.fit")])
Stest2 <- Sapply(berktest2,function(x)coef(summary(x)))
Stest2.1 <- Lapply(berktest1,function(x)predict(x,,se.fit=TRUE)[c("fit","se.fit")])
to.data.frame(Stest1)
to.data.frame(Stest2,as.vars=2)
to.data.frame(Stest2.1)
# Recasting a contingency table
to.data.frame(UCBAdmissions,as.vars="Admit")
}
\keyword{manip}
|
