Surfaces data-quality problems in a cleaned AFP table by reading columns the cleaner already produced (duplicates, blank keys, unreconciled GUIDs, missing/zero coordinates, future onset dates, out-of-range age, negative timeliness intervals, inadequate stool). Checks whose required columns are absent are skipped, so a trimmed input is handled gracefully.
Usage
checks_afp(afp, reference_date = Sys.Date())Arguments
- afp
A cleaned AFP tibble (from
clean_afp()).- reference_date
Date treated as "today" for future-date checks (default
Sys.Date()).
Value
A named list: summary (a tibble with one row per applicable check:
check, domain, severity, n_flagged, description) followed by one
tibble of flagged rows (key columns) per check that found problems. Pass it
to write_checks_excel() to export a workbook.
Examples
afp <- data.frame(
id = c(1, 1), epid = c("A-1", "A-1"),
paralysis_onset_date = c("2024-01-02", "2024-01-02"),
adm0 = "NIGERIA"
)
checks_afp(afp)$summary
#> # A tibble: 3 × 5
#> check domain severity n_flagged description
#> <chr> <chr> <chr> <int> <chr>
#> 1 afp_duplicates AFP warning 2 Duplicate EPID + onset date + admi…
#> 2 afp_no_onset AFP warning 0 AFP cases with no paralysis onset …
#> 3 afp_future_onset AFP warning 0 Onset date later than the run date