xpose.xtras (version 0.2.2)

as_leveler: Level-defining helper functions

Description

as_leveler() is the generic constructor behind every leveler: it tags a character vector x so that set_var_levels() recognizes it as a set of levels rather than a plain formula list. x[1] is mapped to the raw data value .start_index, x[2] to .start_index + 1, and so on. lvl_bin(), lvl_sex(), and lvl_inord() are thin wrappers around as_leveler() for common cases (binary Yes/No, Male/Female sex, and pre-ordered factors, respectively); call as_leveler() directly when defining your own levels, eg more than two categories, a custom starting index, or an unordered custom label set that the built-in wrappers don't cover.

Usage

as_leveler(x, .start_index = 1, .ordered = FALSE)
is_leveler(x)
lvl_bin(x = c("No", "Yes"), .start_index = 0)
lvl_sex()
lvl_inord(x, .start_index = 1, .ordered = TRUE)

Value

Special character vector suitable to be used as leveler

Arguments

x

<character> vector of levels

.start_index

<numeric> starting index for levels

.ordered

<logical> should these levels be treated as an ordered factor (see base::factor) wherever they're consumed (eg val2lvl())?

Examples

# Roll your own leveler for a case the convenience wrappers don't cover,
# eg an unordered, 3-category custom covariate coded 0/1/2 in the data
arm_levels <- as_leveler(c("Placebo", "Low dose", "High dose"), .start_index = 0)
arm_levels
is_leveler(arm_levels)

# The convenience wrappers below are just as_leveler() under the hood, eg
# lvl_bin() is equivalent to as_leveler(c("No", "Yes"), .start_index = 0)
set_var_levels(xpdb_x,
  SEX = lvl_sex(),
  MED1 = lvl_bin(),
  MED2 = lvl_inord(c("n","y"), .start_index = 0)
  )