simple_dis: Simple generation of new variables

Description

Fast simulation of new variables based on univariate distributions

Usage

univariate.dis(puf, data, additional, weights, value = "data", fNA = NA)
conditional.dis(
  puf,
  data,
  additional,
  conditional,
  weights,
  value = "data",
  fNA = NA
)

Arguments

puf: data for which one additional column specified by function argument ‘additional’ is simulated
data: donor data
additional: name of variable to be simulated
weights: sampling weights from data
value: if “data” then the puf including the additional variable is returned, otherwise only the simulated vector.
fNA: only used with missing values if another code as NA should be used
conditional: conditioning variable

Author

Lydia Spies, Matthias Templ

Details

Function uni.distribution: random draws from the weighted univariate distribution of the original data

Function conditional.dis: random draws from the weighted conditional distribution (conditioned on a factor variable)

This are simple functions to produce structural variables, variables that should have the same categories as given ones. For more advanced methods see simCategorical()

Examples

Run this code

## we don't have original data, so let's use eusilc
data(eusilc13puf)
data(eusilcS)
v1 <- univariate.dis(eusilcS, eusilc13puf, additional = "db040",
weights = "rb050", value = "vector")
table(v1)
table(eusilc13puf$db040)
## we don't have original data, so let's use eusilc
##data(eusilc13puf)
##data(eusilcS)
##v1 <- conditional.dis(eusilcS, eusilc13puf, additional = "pb190",
##  conditional = "db040", weights = "rb050")
##table(v1) / sum(table(v1))
##table(eusilc13puf$pb190) / sum(table(eusilc13puf$pb190))