Covariate column names in an environmental data object

Deliberately internal, and a deliberate duplicate of datamatch::covariate_columns(), which is identical. Exporting it would mask datamatch's whenever both packages are attached, for no gain: callers who want it have it from datamatch already. It is not imported from there because datamatch is a suggested rather than a hard dependency — everything in this package operates on the shape datamatch's access functions return, not on datamatch itself, so an object of that shape from any source works. That shape is common to all of them – accessCopernicus(), accessHYCOM(), accessCCMP(), accessFVCOM() and accessERDDAP() – and to anything else that produces one row per location and time step.

Usage

covariate_columns(env_dat)

Arguments

env_dat

an sf POINT object with one row per location and time step, as datamatch's access functions return

Value

character vector of covariate column names

Details

<var>_source is included and <var>_depth is not, matching datamatch::covariate_columns(). A source tag travels with the variable it describes and survives resampling as a categorical; a depth is the model level a derived bottom value was taken from, and the mean of two of those is not the depth any value came from. .datamatch_source is datamatch's internal per-row tag and is never data.