collapseCategories: Collapse multiple categories into a single category.

In StatisticsNZ/dembase: Analysing Cross-Classified Data about Populations

Description

Merge multiple categories within a dimension or dimensions of a DemographicArray object.

Usage

collapseCategories(
  object,
  dimension = NULL,
  old,
  new,
  concordance,
  weights = NULL,
  ...
)

## S4 method for signature 'Counts,ANY,ANY,ANY,missing'
collapseCategories(object, dimension = NULL, old, new, weights = NULL)

## S4 method for signature 'Counts,ANY,missing,missing,OneToOne'
collapseCategories(object, dimension = NULL, concordance, weights = NULL)

## S4 method for signature 'Counts,ANY,missing,missing,ManyToOne'
collapseCategories(object, dimension = NULL, concordance, weights = NULL)

## S4 method for signature 'Values,ANY,ANY,ANY,missing'
collapseCategories(object, dimension, old, new, weights)

## S4 method for signature 'Values,ANY,missing,missing,ManyToOne'
collapseCategories(object, dimension, concordance, weights)

## S4 method for signature 'Values,ANY,missing,missing,OneToOne'
collapseCategories(object, dimension = NULL, concordance, weights = NULL)

Arguments

object	Object of class DemographicArray.
dimension	Character or numeric vector giving the name or index of the dimension where the categories are found - or names or indices if more than one dimension is being changed.
old	Character vector containing the names of the categories to be collapsed.
new	Character vector of length 1 giving the name of the new category formed from the old categories.
concordance	Object of class "Concordance".
weights	Object of class Counts.
...	Not currently used.

Details

The categories to be collapsed, and the categories that result from collapsing can be described through old and new, or through a "ManyToOne" concordance: see below for examples.

Pairs of of dimensions with dimtypes origin and destination or parent and child can be selected by their base name: see the example below. If no dimension is supplied, collapseCategories attempts to modify all dimensions.

weights are required if object has class Values and are prohibited if object has class Counts.

See Also

collapseIntervals, collapseIterations, collapseOrigDest, collapseDimension, dimtypes, dimscales

Examples

titanic <- Counts(Titanic, dimtypes = c(Age = "state"))
collapseCategories(titanic, dimension = "Class",
                   old = c("1st", "2nd", "3rd"),
                   new = "Passengers")
## can also use the dimension index, though the name is clearer
collapseCategories(titanic, dimension = 1,
                   old = c("1st", "2nd", "3rd"),
                   new = "Passengers")
## set up and use a concordance
conc <- data.frame(status = c("1st", "2nd", "3rd", "Crew"),
                   wealth = c("Rich", "Mixed", "Poor", "Mixed"))
conc <- Concordance(conc)
collapseCategories(titanic, dimension = "Class", concordance = conc)

library(demdata)
rates <- Values(VADeaths2)
popn <- Counts(VAPopn)
collapseCategories(rates, dimension = "sex", old = c("Male", "Female"),
                   new = "Combined", weights = popn)

occ <- occupationalStatus
names(dimnames(occ)) <- c("occupation_parent", "occupation_child")
occ <- Counts(occ)
## select a pair of dimensions by specifying a base name
collapseCategories(occ, dimension = "occupation",
                   old = c("2", "5"), new = "9")
## no dimension specified, so changes made to all dimensions
collapseCategories(occ, old = c("2", "5"), new = "9")

StatisticsNZ/dembase documentation built on Dec. 25, 2021, 4:49 p.m.