Enable name-independent key column identification
Source:R/mapper-function-helpers.R
manage_key_colnames.Rd
A handwritten mapper function for consistency tests, such as
grim_map()
, may include arguments named after the key columns in its
input data frame. When such an argument is specified by the user as a
column name of the input data frame, it identifies a differently-named
column as that key column.
Create such functionality in three steps:
Add arguments to your mapper function named after the respective key columns. They should be
NULL
by default; e.g.,x = NULL, n = NULL
.Within the mapper, capture the user input by quoting it using
rlang::enexpr()
. Reassign these values to the argument variables; e.g.,x <- rlang::enexpr(x)
andn <- rlang::enexpr(n)
.For every such argument, call
manage_key_colnames()
and reassign its value to the input data frame variable, adding a short description; e.g.,data <- manage_key_colnames(data, x, "mean/proportion")
anddata <- manage_key_colnames(data, n, "sample size")
.
Arguments
- data
The mapper function's input data frame.
- arg
Symbol. The quoted input variable, captured by
rlang::enexpr()
.- description
String (length 1). Short description of the column in question, to be inserted into an error message.
See also
vignette("consistency-tests-in-depth")
, for context.