% Match between 2 strings

[jlc467]

would be really useful to have a function which calculates the % match between the two given strings. It would tell user how similar the strings are in the form of a %.

[matanox]

There is no single measure of "string similarity" that is intuitively nice for all cases and all minds. This is actually a nice and boring topic in computer science and genomics. You could implement that on your own with the measure of your choice, in your own code. E.g. using levenshtein distance.

[leeoniya]

or Damerau-Levenshtein Distance that treats transposition as 1: http://jsperf.com/damerau-levenshtein-distance. Or http://www.joyofdata.de/blog/comparison-of-string-distance-algorithms/ from the article:

So from a top down perspective a good string metric would consider two strings very close if the first and last letter are matching and the letters in between are just permuted. You don’t have to be a genius to tell from the above given descriptions of the algos that none will perform exceptionally well and the one’s that do are probably just immune to perumtations on a whole – but what the heck – I got curious how the metrics respond to permutations. Okay one further aspect – given that even though human reading seems to be unimpressed by framed permutations ambiguous cases might arise – “ecxept”/”except” and “expcet”/”expect” – then the hamming distance would (maybe) determine the interpretation – which is why I chose it for the coloring in the following plot:

have fun!

[jlc467]

thanks for the response -- have successfully implemented a js lev distance function that suits my needs!