Skip to contents

Classify a column by its variable name

Source: R/02-02-classify-by-name.R
classify_by_name.Rd

Matches a column header against the internal name lexicon (.name_lexicon) and returns the best-matching ontology class. This reaches the metadata-gated classes that value detectors cannot – it reads the name, not the data – so it is a suggestion, weaker than a value-signature match.

Usage

classify_by_name(name, min_conf = 0.5)

Arguments

name

A single variable name (header string).

min_conf

Minimum rule confidence for the top match to be reported as the guess; below it, guess is NA.

Value

A list with:

guess

Best-matching class name, or NA.

ontology

Named vector c(data_type, data_subtype, data_class, data_format); data_format is always NA (a name implies no format).

confidence

The matched rule's confidence.

candidates

Data frame of all matching rules, ranked by confidence.

normalized

The normalized header actually matched against.

See also

guess_column(), guess_data_type()

Examples

classify_by_name("hourly_wage")$guess          # currency
#> [1] "currency"
classify_by_name("respondent_weight")$guess    # weight
#> [1] "weight"
classify_by_name("is_active")$guess            # indicator
#> [1] "indicator"