Skip to contents

Append new rows to metadata/variable.csv, safely

Source: R/registry.R
register_variables.Rd

Mirrors register_measurement_types(): append-only (edit the CSV by hand to change an existing row), refuses a duplicate variable key, writes with na = "".

Usage

register_variables(new_vars, path, measurement_type_path = NULL, quiet = FALSE)

Arguments

new_vars

data.frame with a variable column and any of label, description, units, nerc_p01, category, is_unified. Columns absent from the registry are dropped with a warning.

path

path to metadata/variable.csv

measurement_type_path

path to metadata/measurement_type.csv, used only to check nerc_p01 agreement against member series (rows whose variable equals the new row's variable). NULL skips the check (not recommended — nothing else catches this).

quiet

suppress the "added N variable(s)" message

Value

The full updated registry.

Details

It additionally refuses a row whose nerc_p01 is set but disagrees with a member series' own nerc_p01 in measurement_type.csv — the whole point of the crosswalk key is that agreement can be asserted once, and a silent mismatch would ship the wrong concept URI to an OBIS/DwC export for every series sharing the key but one. It does not refuse a key whose member series disagree with each OTHER (two series under the same variable that carry two different NERC concepts) — that is a real fact about the data (sigma_theta: the bottle's own computation is SIGTEQ01, the CTD's is SIGTPR01 — different P01 concepts, found WS-M1 2026-09-10), and the correct way to record it is to leave the row's own nerc_p01 empty, exactly as an id is left empty anywhere else in these registries when no single concept says exactly what a row is (metadata-registries skill). Passing a non-empty nerc_p01 for such a key is refused, because it would silently pick a side.

See also

check_variable_registry(), the standing invariant this enforces at write time.

Examples

if (FALSE) { # \dontrun{
register_variables(
  data.frame(variable = "temperature", label = "Temperature",
             units = "degC",
             nerc_p01 = "http://vocab.nerc.ac.uk/collection/P01/current/TEMPPR01/",
             category = "Physical Oceanography", is_unified = TRUE),
  here::here("metadata/variable.csv"),
  here::here("metadata/measurement_type.csv"))
} # }