Dear community,

I have merged two datasets based on a unique companyid with the reclink command.


reclink company_lender_id using "C:\Users\huett\OneDrive\Dokumente\SS_19\Kaserer\s nydicated_loans\lenders - unique.dta", idmaster(idmaster
> ) idusing(idlender) gen(score2) _merge(mergedata2) minscore(.99)

905 perfect matches found


Added: idlender= identifier from C:\Users\huett\OneDrive\Dokumente\SS_19\Kaserer\sn ydicated_loans\lenders - unique.dta score2 = matching s
> core
Observations: Master N = 906 C:\Users\huett\OneDrive\Dokumente\SS_19\Kaserer\sn ydicated_loans\lenders - unique.dta N= 19920
Unique Master Cases: matched = 896 (exact = 905), unmatched = 10



After having run the command several times I noticed that there are 10 observations in total which stata refuses to match although the variable which is used for matching is exactly the same in both data sets (namely the unique identifier). I was wondering whether anyone has encountered a similar Problem with reclink so far. What strikes me is that it is the same 10 observations every time that remains unmatched whereas the algorithm works perfectly for the rest of the ~ 1000 observations. I have checked that there are no Special characters such as äüö etc.

What am I missing?

Thanks a lot!

Marie