Add a binary Hispanic-origin column from hspanicr

Adds a hispanic_origin column ("hispanic" / "non_hispanic" / "unknown" / NA) derived from hspanicr and the data year, so a death frame can be joined to Hispanic-stratified population denominators via add_pop_counts. This is the population-join counterpart of add_hspanicr_column (which backfills the raw hspanicr field): Hispanic origin is ragged across years – absent before 1989, reserved in 2021, and recoded from 9 to 14 categories in 2022 – so, like add_hspanicr_column(), this reads the canonical year and tolerates a missing/NA hspanicr (yielding NA origin for those rows).

Usage

add_hispanic_origin(df)

Arguments

df

an MCOD dataframe with hspanicr (or none, treated as NA) and a year or datayear column.

Value

df with an added hispanic_origin character column.

Details

Year is read per row: year (1996+) where present, otherwise the two-digit datayear (1979-1995, normalized to four digits). When a frame carries both columns – e.g. a bind_rows() of pre-1996 and 1996+ chunks, where each era's column is NA outside its own rows – the two are coalesced per row, so every row is labeled by its own data year. It errors if neither column is present.

See also

categorize_hispanic_origin (the vectorized recode); add_hspanicr_column (backfills the raw hspanicr field); add_pop_counts for the Hispanic-stratified population join.

Examples

df <- data.frame(year = 2019, hspanicr = c(1, 6, 9))
add_hispanic_origin(df)
#>   year hspanicr hispanic_origin
#> 1 2019        1        hispanic
#> 2 2019        6    non_hispanic
#> 3 2019        9         unknown