Recode another agency's quality flags to the package's

The comparisons of this family read Statistics Canada's quality flags: E (use with caution), and F, x, .., ... (no value: too unreliable, suppressed, not available, not applicable). Other agencies flag otherwise: the Australian Bureau of Statistics np (not published) and * or ** (relative standard error over 25% or 50%), Stats NZ S (suppressed), the US Census Bureau (X). This maps them to the package's vocabulary, so suppressed cells are never compared and uncertain ones are marked.

Usage

qltyrcd(x, suppr = c("F", "x", "..", "..."), caution = "E")

Arguments

x

The flags, one per row.

suppr

Flags meaning the value must not be used (it becomes "x").

caution

Flags meaning the value may be used with caution (it becomes "E").

Value

A character vector of "ok", "E" and "x"; empty and missing flags are "ok". x must be character (or a factor); a flag in neither suppr nor caution (other than "ok") becomes "ok" and is named in a warning.

See also

limfrom, indgap.

Examples

# flags as an Australian table might give them
qltyrcd(c("", "np", "*", "**", NA), suppr = c("np", "**"), caution = "*")
#> [1] "ok" "x"  "E"  "x"  "ok"