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
|
\name{contract}
\alias{contract}
\alias{contract.data.frame}
\alias{contract.data.set}
\title{Contract data into pattern-frequency format}
\description{
\code{contract()} contracts data into pattern-frequency format, similar
to a contatenation of \code{table()} (or \code{xtabs}) and
\code{as.data.frame()}. Yet it uses much less memory if patterns
are sparse, because it does not create rows for patterns that do not occur.
}
\usage{
contract(x,\dots)
\method{contract}{data.frame}(x,by=NULL, weights=NULL,name="Freq",
force.name=FALSE,sort=FALSE,drop.na=TRUE,\dots)
\method{contract}{data.set}(x,by=NULL, weights=NULL,name="Freq",
force.name=FALSE,sort=FALSE,drop.na=TRUE,\dots)
}
\arguments{
\item{x}{
an object of class \code{"data.frame"} or \code{"data.set"}.
}
\item{by}{the formula or a vector of variable names (quoted or not quoted).
Specifies the patterns (and optionally weights).
If \code{by} is a formula, then the right-hand side specifies the
variables the value patterns of which are counted.
If the left-hand side of the formula is (the name of) a numeric
vector, its values are used as weights (in which case the
\code{weights} argument will be ignored.) If the left-hand side of
the formula is (the name of) a factor, counts are computed in
separate columns for each of its levels.
}
\item{weights}{a numeric vector of weights or \code{NULL}.}
\item{name}{a character string, the name of the variable that
containts the frequency counts of the value patterns.}
\item{force.name}{a logical value, defaults to \code{FALSE}. If
\code{TRUE} and the left-hand side of \code{by} formula is a factor,
the names of the columns with the counts are combinations of the
labels of the factor levels and the argument of \code{name}; if
\code{FALSE}, the column names are created from the labels of the
factor levels only.
}
\item{sort}{a logical value, defaults to \code{FALSE}. If \code{TRUE},
the resulting data set is sorted by the variables that define the
patterns. If \code{FALSE}, the row of the resulting data frame or
data set are ordered according to the occurrence of the patterns.
}
\item{drop.na}{a logical value, defaults to \code{TRUE}. If
\code{FALSE}, patterns that involve \code{NA} are included in the
resulting data frame or data set.}
\item{\dots}{further arguments, passed to methods or ignored.}
}
\value{
If \code{x} is a data fame, the value of \code{contract()} is also a
data frame. If it is a \code{"data.set"} object, the result is also a
\code{"data.set"} object.
}
\examples{
iris_ <- sample(iris,size=nrow(iris),replace=TRUE)
w <- rep(1,nrow(iris_))
contract(iris[4:5])
contract(iris[4:5],sort=TRUE)
contract(iris[4:5],weights=w,sort=TRUE)
contract(iris,by=c(Petal.Width,Species),sort=TRUE)
contract(iris,by=~Petal.Width+Species)
contract(iris,by=w~Species)
library(MASS)
contract(housing,
by=Sat~Infl+Type+Cont,
weights=Freq)
contract(housing,
by=Sat~Infl+Type+Cont,
weights=Freq,
name="housing",force.name=TRUE
)
}
|
