How to ensure authors have a unique name - r

Question

I extracted publication data from Microsoft Academic. Unfortunately, some authors have different versions of their names, e.g.
names <- data.frame(publication_id = c("1", "2", "3", "4", "5"), author = c("D Smith", "D J Smith", "David Smith", "Enrique Salvador", "E J Salvador"), affiliation = c("UCT", "UCT", "UCT", "UTAS", "UTAS")
I want authors to have a unique name. Using the above example, I want to get a result that looks like this:
names <- data.frame(publication_id = c("1", "2", "3", "4", "5"), author = c("D. Smith", "D Smith", "D Smith", "E Salvador", "E Salvador"), affiliation = c("UCT", "UCT", "UCT", "UTAS", "UTAS")
I am dealing with 1000s of author names so using something like:
mutate(author = case_when(author == "D.J. Smith" ~ "D. Smith", author == "David Smith" ~ "D.Smith",
is impractical. I would appreciate any ideas/solutions. Thanks in advance.

Answer 1

I think the most likely thing to work would be to take the first letter of the string and the last name and put them together:
names <- data.frame(publication_id = c("1", "2", "3", "4", "5"), author = c("D Smith", "D J Smith", "David Smith", "Enrique Salvador", "E J Salvador"), affiliation = c("UCT", "UCT", "UCT", "UTAS", "UTAS"))
names$new_author <- paste0(substr(names$author, 1, 1),
" ",
gsub('.*\\s(\\w*)$', '\\1', names$author))
names
# publication_id author affiliation new_author
# 1 1 D Smith UCT D Smith
# 2 2 D J Smith UCT D Smith
# 3 3 David Smith UCT D Smith
# 4 4 Enrique Salvador UTAS E Salvador
# 5 5 E J Salvador UTAS E Salvador
That said, it assumes that there is only one D Smith at UCT which may be an untenable assumption.