Statistical-Functions: Existing statistical functions.

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

Description

Existing statistical functions give the same result when used with a DemographicArray array as they would with an ordinary array - except when the demographic array has a dimension with dimtype "iteration" or "quantile". When there is an "iteration" dimension, the calculations are repeated once for each iteration, so that the return value is also a demographic array with an "iteration" dimension.

Usage

## S4 method for signature 'DemographicArrayOrNumeric'
Summary(x, ..., na.rm = FALSE)

## S4 method for signature 'DemographicArray'
mad(
  x,
  center = median(x),
  constant = 1.4826,
  na.rm = FALSE,
  low = FALSE,
  high = FALSE
)

## S4 method for signature 'DemographicArray'
mean(x, ...)

## S4 method for signature 'DemographicArray'
median(x, na.rm = FALSE, ...)

Arguments

x	Typically an object of class DemographicArray.
...	Other arguments.
na.rm	Logical. Whether to remove missing values before calculating.
center	Function used to calculate central value. Defaults to median.
constant	See documentation for mad.
low	See documentation for mad.
high	See documentation for mad.

Details

When called on an ordinary array, the functions all return a single number, except for range, which returns 2. When called on a demographic array with an "iteration" dimension, range returns a demographic array with a dimension called "range", in addition to the "iteration" dimension.

When a demographic array has a dimension with dimtype "quantile", applying statistical functions typically does not make sense, and an error is raised. It is generally better to work with the original array, with the complete set of iterations, and to reduce to quantiles once the calculations are complete.

Value

Numeric, if x does not contain an "iteration" dimension, and Counts or Values if it does.

See Also

Iteration and quantile dimensions are described in the documentation for dimtypes. To remove iteration dimensions (by summarising across them) use collapseIterations.

Examples

## no iteration dimension
x1 <- Counts(array(1:4,
                   dim = 4,
                   dimnames = list(region = c("A", "B", "C", "D"))))
## with iteration dimension
x2 <- Counts(array(c(1:4, 0:3, 2:5),
                   dim = c(4, 3),
                   dimnames = list(region = c("A", "B", "C", "D"),
                                   iteration = 1:3)))
x1
x2
sum(x1)
sum(x2)
range(x1)
range(x2)

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