List affiliation strings that matched no institution
Source:R/affiliation.R
sm_affiliation_unmatched.RdAudits an affiliation dictionary. After sm_affiliation_match(), it lists
the most frequent raw_affiliation strings of the authorships that no
signal (name pattern, email domain or postcode) assigned to an
institution, so the dictionary can be extended where that matters most.
Usage
sm_affiliation_unmatched(corpus, n = 25L, call = rlang::caller_env())Arguments
- corpus
An
sm_corpuspreviously passed throughsm_affiliation_match()(so itsauthorshipscarryinstitution_match).- n
Maximum number of strings to return: a positive whole number, or
Inffor all of them. Default 25.- call
Caller environment for error reporting.
Value
A tibble with columns raw_affiliation (character),
n_authorships (integer: unmatched authorships with that string) and
n_works (integer: distinct works among them), at most n rows,
sorted by n_authorships descending and then by raw_affiliation in
code-point order. Authorships whose raw_affiliation is NA or blank
are not listed. Type-stable: a 0-row tibble when every affiliation
matched, and a 0-row tibble with a warning when the corpus has not been
matched.
Examples
fix <- readRDS(sm_example_files("example_realuse_corpus.rds"))
dict <- readr::read_csv(
sm_example_files("example_realuse_affiliation_dict.csv"),
col_types = "ccc"
)
matched <- sm_affiliation_match(fix, dict)
#> ✔ Affiliation matching flagged 427 authorships across 4 institutions.
#> ℹ By signal: name_token: 427. See `sm_affiliation_summary()` for the full
#> breakdown.
sm_affiliation_unmatched(matched, n = 5)
#> # A tibble: 5 × 3
#> raw_affiliation n_authorships n_works
#> <chr> <int> <int>
#> 1 Hilltop Univ Med Ctr, Dept Surg, NL-9999 ZZ Hilltop, Ne… 103 52
#> 2 Univ Northtown, Inst Clin Chem, Northtown, Germany. 46 35
#> 3 Alpenau Cantonal Hosp, Dept Infect Dis, CH-9999 Alpenau… 42 33
#> 4 Univ Northtown, Dept Microbiol, Northtown, Germany. 33 26
#> 5 Univ Northtown, Dept Trauma Surg, Northtown, Germany. 31 26