SingletonUniqueContributor {GaussSuppression}

Unique contributor singleton function

Description

Function for GaussSuppressionFromData

Usage

SingletonUniqueContributor(
  data,
  freqVar = NULL,
  nUniqueVar = NULL,
  charVar = NULL,
  removeCodes = character(0),
  integerSingleton = length(charVar) > 0,
  x,
  primary = integer(0),
  whenPrimaryMatters = warning,
  whenNoVar = TRUE,
  specialMultiple = TRUE,
  ...
)

SingletonUniqueContributor0(data, numVar, dominanceVar = NULL, ...)

Arguments

data

Input data, possibly pre-aggregated within GaussSuppressionFromData

freqVar

A single variable holding counts (input to GaussSuppressionFromData)

nUniqueVar

A single variable holding the number of unique contributors.

charVar

Variable with contributor codes.

removeCodes

Vector, list or data frame of codes considered non-singletons. Single element lists and single column data frames behave just like vectors. In other cases, charVar-names must be used. With empty charVar a vector of row indices is assumed and conversion to integer is performed. See examples.

integerSingleton

Integer output when TRUE. See details.

x

ModelMatrix generated by parent function

primary

Vector (integer or logical) specifying primary suppressed cells. It will be ensured that any non-suppressed inner cell is not considered a singleton.

whenPrimaryMatters

Function to be called when primary caused non-singleton. Supply NULL to do nothing.

whenNoVar

When TRUE, and without nUniqueVar and freqVar in input, all cells will be marked as singletons.

specialMultiple

When TRUE, and when ⁠integerSingleton &⁠ length(charVar) > 1 ⁠& length(nUniqueVar)⁠, a special method is used. By re-coding to single charVar and by re-calculating nUnique. To be unique (nUnique=1), uniqueness is only required for a single charvar. Otherwise, the charvar combination must be unique.

...

Unused parameters

numVar

vector containing numeric values in the data set

dominanceVar

When specified, dominanceVar is used in place of numVar. Specifying dominanceVar is beneficial for avoiding warnings when there are multiple numVar variables. Typically, dominanceVar will be one of the variables already included in numVar.

Details

This function marks input cells as singletons according to ones in data[[nUniqueVar]], if available, and otherwise according to data[[freqVar]]. The output vector can be logical or integer. When, integer, singletons are given as positive values. Their unique values represent the unique values/combinations of data[[charVar]].

Value

logical or integer vector

Note

SingletonUniqueContributor0 is a special version that produces singleton as a two-element list. See GaussSuppression and SuppressDominantCells.

Examples

S <- function(data, ...) {
  cbind(data, singleton = SingletonUniqueContributor(data, ...))
}
d2 <- SSBtoolsData("d2")
d <- d2[d2$freq < 5, ]
d$nUnique <- round((5 - d$freq)/3)
d$freq <- round(d$freq/2)
d[7:8, 2:4] <- NA
rownames(d) <- NULL

S(d, freqVar = "freq", integerSingleton = FALSE)
S(d, freqVar = "freq", nUniqueVar = "nUnique", integerSingleton = TRUE, charVar = "main_income")
S(d, nUniqueVar = "nUnique", integerSingleton = TRUE, charVar = c("main_income", "k_group"))
S(d, freqVar = "freq", nUniqueVar = "nUnique", integerSingleton = FALSE, 
  charVar = "main_income", removeCodes = "other")
S(d, nUniqueVar = "nUnique", integerSingleton = FALSE, charVar = c("main_income", "k_group"), 
  removeCodes = c("other", "400"))
S(d, nUniqueVar = "nUnique", integerSingleton = FALSE, charVar = c("main_income", "k_group"), 
  removeCodes = data.frame(anyname = c("other", "400")))
S(d, nUniqueVar = "nUnique", integerSingleton = FALSE, charVar = c("main_income", "k_group"), 
  removeCodes = list(main_income = c("other", "pensions"), k_group = "300"))
S(d, nUniqueVar = "nUnique", integerSingleton = FALSE, charVar = c("main_income", "k_group"), 
  removeCodes = data.frame(main_income = "other", k_group = "400"))
S(d, nUniqueVar = "nUnique", integerSingleton = FALSE, removeCodes = 1:5)

x <- SSBtools::ModelMatrix(d, hierarchies = list(region = "Total"))
which(colSums(x) == 1)
which(rowSums(x[, colSums(x) == 1]) > 0)
# columns 2, 3, 4, 5, 7 correspond to inner cells: rows 3, 4, 5, 6, 8 
# with 2:4 not primary rows 3:5 are forced non-singleton
S(d, freqVar = "freq", nUniqueVar = "nUnique", integerSingleton = FALSE, x = x, primary = 5:8)