Skip to contents

Standardise column names onto the NARWC handbook's

Source: R/read-narwc.R
standardize_narwc_columns.Rd

Renames a data frame's columns to their canonical NARWC names, matching case-insensitively and ignoring separators. This is the step read_narwc() does first, exported on its own so that other packages working with NARWC survey data can use the same vocabulary rather than maintain a second one.

Usage

standardize_narwc_columns(dat, quiet = FALSE, prefer_source = TRUE)

Arguments

dat

A data frame.

quiet

Suppress the report of inferred matches. Default FALSE.

prefer_source

Let a Trk* GPS track column take precedence over a canonical column of the same name. Default TRUE. See read_narwc().

Value

dat with columns renamed where a match was found.

Why this is shared

Every pipeline that reads real survey exports hits the same wall: the files say Event, event_no, or Event No. and the code wants EVENTNO. Solving it once per package means the vocabularies drift, and a column recognised in one place is silently dropped in another. The alias table here is the merge of two such attempts — distsamp's and msomgom's — and this package exists so that it is the only one that grows.

What it will not do

Guess by edit distance. EVENTN0 with a zero stays EVENTN0, because a column that is nearly a name is not that name. It also never renames onto a canonical column that already exists, so a correctly named column always wins, and it warns when two columns could both be one canonical name.

Policy stays with the caller. This renames columns and nothing else: it does not fill defaults, drop records, or coerce types. read_narwc() does those, and packages with different needs keep their own on top. The standing example is ALT: msomgom defaults a missing altitude to a nominal survey height, which is reasonable for occupancy but would be wrong in distsamp, where ALT feeds a right-angle distance and a fabricated altitude gives a fabricated distance. Neither default belongs here.

See also

read_narwc(), narwc_schema()

Examples

raw <- data.frame(Event = 1, Lat = 43, Long = -69, Sea_State = 3,
                  check.names = FALSE)
names(standardize_narwc_columns(raw, quiet = TRUE))
#> [1] "EVENTNO"   "LATITUDE"  "LONGITUDE" "BEAUFORT"