Skip to contents

Assert that conditions hold

Source: R/assert.R
assert.Rd

Verify that every row satisfies all specified conditions, like Stata's assert. Stop with an error if any condition fails; otherwise, return the original data invisibly.

Usage

assert(.data, ..., .by = NULL)

Arguments

.data

A data frame, tibble, or data.table.

...

Logical conditions using columns of .data. Every condition must be TRUE for every row.

.by

<tidy-select> Optional grouping variables. Summaries such as mean(x) are evaluated within each group.

Value

.data, unchanged and invisibly, if all conditions pass.

Details

Both FALSE and NA count as failures. Errors report the number of failing rows for each condition, including missing results, and the first five failing row numbers. A failed group-level condition counts every row in that group.

Unlike Stata, x > 0 fails when x is missing. Use is.na(x) | x > 0 to allow missing values.

See also

Other data checking functions: is_first(), keys, stata_merge()

Examples

scores <- data.frame(
  class = c("A", "A", "B", "B"),
  score = c(80, 90, 55, NA_real_)
)

# Pass: every nonmissing score is at least 50
assert(scores, is.na(score) | score >= 50)

# Fail: row 3 is below 60, and row 4 is missing
try(assert(scores, score >= 60))
#> Error in assert(scores, score >= 60) : Assertion failed:
#> * `score >= 60`: 2 contradictions (1 missing) in 4 rows; rows 3, 4