data.set manipulation memisc 0.99.26.3

Manipulation of Data Sets

Description

Like data frames, data.set objects have subset, unique, cbind, rbind, merge methods defined for them.

The semantics are basically the same as the methods defined for data frames in the base package, with the only difference that the return values are data.set objects. In fact, the methods described here are front-ends to the corresponding methods for data frames, which are constructed such that the “extra” information attached to variables within data.set objects, that is, to item objects.

Usage

## S4 method for signature 'data.set'
subset(x, ...)

## S4 method for signature 'data.set'
unique(x, incomparables = FALSE, ...)

## S3 method for class 'data.set'
cbind(..., deparse.level = 1)

## S3 method for class 'data.set'
rbind(..., deparse.level = 1)

## S4 method for signature 'data.set,data.set'
merge(x,y, ...)

## S4 method for signature 'data.set,data.frame'
merge(x,y, ...)

## S4 method for signature 'data.frame,data.set'
merge(x,y, ...)

Arguments

x, y

data.set objects. On of the arguments to merge may also be an object coercable into a data frame and the result still is a data.set object.

...

for subset: a logical vector of the same length as the number of rows of the data.set and, optionally, a vector of variable names (tagged as select); for unique: further arguments, ignored; for cbind, rbind: objects coercable into data frames, with at least one being a data.set object; for merge: further arguments such as arguments tagged with by, by.x, by.y, etc. that specify the variables by which to merge the data sets of data frames x and y.

incomparables

a vector of values that cannot be compared. See unique.

deparse.level

an argument retained for reasons of compatibility of the default methods of cbind and rbind.

Examples

ds1 <- data.set(
     a = rep(1:3,5),
     b = rep(1:5,each=3)
 )
ds2 <- data.set(
     a = c(3:1,3,3),
     b = 1:5
 )
ds1 <- within(ds1,{
     description(a) <- "Example variable 'a'"
     description(b) <- "Example variable 'b'"
 })
ds2 <- within(ds2,{
     description(a) <- "Example variable 'a'"
     description(b) <- "Example variable 'b'"
 })
str(ds3 <- rbind(ds1,ds2))

Data set with 20 obs. of 2 variables:
 $ a: Itvl. item  num  1 2 3 1 2 3 1 2 3 1 ...
 $ b: Itvl. item  int  1 1 1 2 2 2 3 3 3 4 ...

description(ds3)

a 'Example variable 'a''
b 'Example variable 'b''

ds3 <- within(ds1,{
       c <- a
       d <- b
       description(c) <- "Copy of variable 'a'"
       description(d) <- "Copy of variable 'b'"
       rm(a,b)
   })
str(ds4 <- cbind(ds1,ds3))

Data set with 15 obs. of 4 variables:
 $ ds1.a: Itvl. item  int  1 2 3 1 2 3 1 2 3 1 ...
 $ ds1.b: Itvl. item  int  1 1 1 2 2 2 3 3 3 4 ...
 $ ds3.c: Itvl. item  int  1 2 3 1 2 3 1 2 3 1 ...
 $ ds3.d: Itvl. item  int  1 1 1 2 2 2 3 3 3 4 ...

description(ds4)

ds1.a 'Example variable 'a''
ds1.b 'Example variable 'b''
ds3.c 'Copy of variable 'a''
ds3.d 'Copy of variable 'b''

ds5 <- data.set(
       c = 1:3,
       d = c(1,1,2)
       )
ds5 <- within(ds5,{
     description(c) <- "Example variable 'c'"
     description(d) <- "Example variable 'd'"
 })
str(ds6 <- merge(ds1,ds5,by.x="a",by.y="c"))

Data set with 15 obs. of 3 variables:
 $ a: Itvl. item  int  1 1 1 1 1 2 2 2 2 2 ...
 $ b: Itvl. item  int  1 4 3 2 5 1 4 3 2 5 ...
 $ d: Itvl. item  num  1 1 1 1 1 1 1 1 1 1 ...

# Note that the attributes of the left-hand variables
# have priority.
description(ds6)

a 'Example variable 'a''
b 'Example variable 'b''
d 'Example variable 'd''