linux - Merging through fuzzy matching of variables in R

Question

I have two dataframes (x & y) where the IDs are student_name, father_name and mother_name. Because of typographical errors ("n" instead of "m", random white spaces, etc.), I have about 60% of values which are not aligning, though I can eyeball the data and see they should. Is there a way to reduce the level of non-match somehow so that manually editing because at least feasible? The dataframes are have about 700K observations.

R would be best. I know a little bit of python, and some basic unix tools. P.S. I read up on agrep(), but don't understand how that can work on actual datasets, especially when the match is over more than one variable.

---

update (data for posted bounty):

Here are two example data frames, sites_a and sites_b. They could be matched on the numeric columns lat and lon as well as on the sitename column. It would be useful to know how this could be done on a) just lat + lon, b) sitename or c) both.

you can source the file test_sites.R which is posted as a gist.

Ideally the answer would end with

merge(sites_a, sites_b, by = **magic**)

Answer 1

The agrep function (part of base R), which does approximate string matching using the Levenshtein edit distance is probably worth trying. Without knowing what your data looks like, I can't really suggest a working solution. But this is a suggestion... It records matches in a separate list (if there are multiple equally good matches, then these are recorded as well). Let's say that your data.frame is called df:

l <- vector('list',nrow(df))
matches <- list(mother = l,father = l)
for(i in 1:nrow(df)){
  father_id <- with(df,which(student_name[i] == father_name))
  if(length(father_id) == 1){
    matches[['father']][[i]] <- father_id
  } else {
    old_father_id <- NULL
    ## try to find the total                                                                                                                                 
    for(m in 10:1){ ## m is the maximum distance                                                                                                             
      father_id <- with(df,agrep(student_name[i],father_name,max.dist = m))
      if(length(father_id) == 1 || m == 1){
        ## if we find a unique match or if we are in our last round, then stop                                                                               
        matches[['father']][[i]] <- father_id
        break
      } else if(length(father_id) == 0 && length(old_father_id) > 0) {
        ## if we can't do better than multiple matches, then record them anyway                                                                              
        matches[['father']][[i]] <- old_father_id
        break
      } else if(length(father_id) == 0 && length(old_father_id) == 0) {
        ## if the nearest match is more than 10 different from the current pattern, then stop                                                                
        break
      }
    }
  }
}

The code for the mother_name would be basically the same. You could even put them together in a loop, but this example is just for the purpose of illustration.