anonymise: Anonymise factor columns across datasets in a consistent way

In simplanonym: Consistent Anonymisation Across Datasets

Anonymise factor columns across datasets in a consistent way

Description

'anonymise()' is a useful function for anonymising factor columns across different datasets using consistent anonymised levels. In other words, if the same factor level appears in more than one dataset, then 'anonymise()' will use the same anonymous factor for that level.

Usage

anonymise(data_list, prefix = "", return_original_levels = FALSE)

Arguments

data_list	A list of data frames or tibbles.
prefix	A character prefix to insert in front of the random labels.
return_original_levels	Whether or not the resulting list should also include the original, non-anonymised levels. Default: FALSE.

Value

A list containing the original data, but with consistently anonymised factors

Examples

library(simplanonym)

rand_tbl_1 <- vroom::gen_tbl(10, 4, col_types = "fffd")
rand_tbl_2 <- vroom::gen_tbl(10, 2, col_types = "fd")
rand_tbl_2$X3 <- rand_tbl_1$X3

# note:
# * rand_tbl_1 and rand_tbl_2 share three column names,
#   of which X2 is a factor in one but not the other.
# * X1 factors do not overlap, but their anonymisation
#   should still be consistent (ie, different levels should
#'#   have their own unique anonymised factors).
# * For X3, the anonymised factors should consider the levels
#   at both `rand_tbl_1$X3` and `rand_tbl_2$X3`.

data_list <- list(rand_tbl_1, rand_tbl_2)
data_list

data_list |> anonymise(return_original_levels = TRUE)

simplanonym documentation built on Feb. 16, 2023, 6:23 p.m.