Match chemical names against WQX Characteristic Name dictionary
Source:R/wqx_matching.R
match_wqx.RdRuns a three-tier lookup: (1) exact canonical, (2) alias crosswalk, (3) Jaro-Winkler fuzzy against canonical names only. Returns a tibble with one row per input name.
Arguments
- names
Character vector of analyte names to match
- dictionary
Tibble from load_wqx_dictionary() with columns: name, canonical_name, type, and optionally cas_number. CAS evidence comes only from canonical entries.
- threshold
Numeric similarity threshold for fuzzy acceptance (default 0.85). Internally converted to JW distance cutoff: distance <= (1 - threshold).
- verbose
Logical. If TRUE, emits per-name cli output (default FALSE).
Value
Tibble with columns: input_name, wqx_name, match_tier, match_distance, alias_type,
wqx_cas, wqx_cas_status, wqx_cas_raw, and wqx_cas_provenance.
wqx_cas contains a single format/checksum-valid canonical-entry CAS, never
an accepted input identity. Status is valid, missing, invalid, or ambiguous.
Distinct nonempty raw canonical values are retained in wqx_cas_raw, separated
by |. Conflicting values yield ambiguous status and missing wqx_cas.
Provenance records canonical: followed by the normalized canonical key.
Unmatched inputs have missing CAS status and no provenance.