Coerce, derive and translate the subject columns of a metadata table

The half of load_patient_table() that operates on an already-selected data.frame: numeric coercion, placeholder rejection, age derivation from date of birth, and dictionary translation of free-text values.

Usage

normalise_patient_columns(
  out,
  value_map = default_value_map(),
  reference_date = Sys.Date()
)

Arguments

out

data.frame whose columns already carry canonical English names.

value_map

The value map. Default default_value_map().

reference_date

The reference date. Default Sys.Date().

Value

out with the same columns, coerced and translated, carrying an untranslated attribute.

Details

WHY IT IS A SEPARATE FUNCTION. Subject attributes now arrive by two routes: a standalone patient table, and the subject columns of a unified sample sheet (see read_samplesheet()). Both must apply the same coercions, or the same study analysed through the two routes would produce different ages and different sex labels. Sharing the implementation is what makes the two routes equivalent rather than merely similar.