Apply a list of validation rules to a data frame — ks_check

Evaluates each rule formula against data and records a message for every row that violates a rule. Two modes are supported:

"first": Each row gets the message of the first rule it fails; subsequent rules are not evaluated for that row.
"all": All rules are evaluated for every row; the row receives a list of all messages for rules it fails.

Usage

ks_check_rules(data, rules, mode = c("first", "all"))

Arguments

data

A data frame.

rules

A named list of rule objects, each with two fields:

expr: A one-sided formula whose RHS is a logical expression (evaluated with rlang::eval_tidy()) that is TRUE for failing rows.
msg: Character scalar. Message written to failing rows.

mode

Character. Either "first" (default) or "all".

Value

data with an additional list-column check_msgs. In "first" mode each element is a single string ("OK" for passing rows). In "all" mode each element is a character vector of all violated messages (or "OK").

Examples

if (FALSE) { # \dontrun{
rules <- list(
  list(expr = ~ AGE < 0, msg = "Negative age"),
  list(expr = ~ is.na(SEX), msg = "Missing sex")
)
ks_check_rules(adsl, rules, mode = "all")
} # }