R function for a function to be repeated based on column values

Question

I have a large data set for which I need to do string matching. I have got some very useful posts from this site and referring them I have created a function to do the strin

Answer 1

While you can probably use an apply function to repeat over separate data files of different regions, here is a fuzzyjoin solution based on my answer to your previous question.

It looks for the best stringdist match for Address and the AreaCode must match exactly (==). I also specified year2 had to be >= year1, just for demonstration.

Finally, I used dplyr::group_by and dplyr::top_n to get the minimum distance matches and I had to assume what to do in matching ties (picked matches with largest year2). You can also use slice_min which replaces the older top_n and if the original order is important and not alphabetical, use mutate(rank = row_number(dist)) %>% filter(rank == 1)

Data:

Address1 <- c("786, GALI NO 5, XYZ","rambo, 45, strret 4, atlast, pqr","23/4, 23RD FLOOR, STREET 2, ABC-E, PQR","45-B, GALI NO5, XYZ","HECTIC, 99 STREET, PQR")
AREACODE <- c('10','10','14','20','30')
Year1 <- c(2001:2005)

Address2 <- c("abc, pqr, xyz","786, GALI NO 4 XYZ","45B, GALI NO 5, XYZ","del, 546, strret2, towards east, pqr","23/4, STREET 2, PQR","abc, pqr, xyz","786, GALI NO 4 XYZ","45B, GALI NO 5, XYZ","del, 546, strret2, towards east, pqr","23/4, STREET 2, PQR")
Year2 <- c(2001:2010)
AREA_CODE <- c('10','10','10','20','30','40','50','61','64', '99')

data1 <- data.table(Address1, Year1, AREACODE)
data2 <- data.table(Address2, Year2, AREA_CODE)
data2[, unique_id := sprintf("%06d", 1:nrow(data2))]

Solution:

library(fuzzyjoin, quietly = TRUE); library(dplyr, quietly = TRUE)

# First, need to define match_fun_stringdist 
# Code from stringdist_join from https://github.com/dgrtwo/fuzzyjoin
match_fun_stringdist <- function(v1, v2) {
  
  # Can't pass these parameters in from fuzzy_join because of multiple incompatible match_funs, so I set them here.
  ignore_case = FALSE
  method = "dl"
  max_dist = 99
  distance_col = "dist"
  
  if (ignore_case) {
    v1 <- stringr::str_to_lower(v1)
    v2 <- stringr::str_to_lower(v2)
  }
  
  # shortcut for Levenshtein-like methods: if the difference in
  # string length is greater than the maximum string distance, the
  # edit distance must be at least that large
  
  # length is much faster to compute than string distance
  if (method %in% c("osa", "lv", "dl")) {
    length_diff <- abs(stringr::str_length(v1) - stringr::str_length(v2))
    include <- length_diff <= max_dist
    
    dists <- rep(NA, length(v1))
    
    dists[include] <- stringdist::stringdist(v1[include], v2[include], method = method)
  } else {
    # have to compute them all
    dists <- stringdist::stringdist(v1, v2, method = method)
  }
  ret <- dplyr::data_frame(include = (dists <= max_dist))
  if (!is.null(distance_col)) {
    ret[[distance_col]] <- dists
  }
  ret
}

# Finally, call fuzzy_join
fuzzy_join(data1, data2, 
           by = list(x = c("Address1", "AREACODE", "Year1"), y = c("Address2", "AREA_CODE", "Year2")), 
           match_fun = list(match_fun_stringdist, `==`, `<=`),
           mode = "left"
           ) %>%
  group_by(Address1, Year1, AREACODE) %>%
  top_n(1, -Address1.dist) %>%
  top_n(1, Year2) %>%
  select(unique_id, Address1.dist, everything())