>>> import Levenshtein as levd
>>> a='MOZILLA'
>>> b='MUSIAL'
>>> levd.distance(a, b)
4
This makes sense if one allows transpositions (the errors are O-U, Z-S, ins L, swap LA), but this is not documented anywhere (AFAIK) and means this package is actually implementing Damerau-Levenshtein distance?
This makes sense if one allows transpositions (the errors are O-U, Z-S, ins L, swap LA), but this is not documented anywhere (AFAIK) and means this package is actually implementing Damerau-Levenshtein distance?