Is there a way to compare two strings for likeness?

Question

I have an issue where i need to link certain sample names to eachother, the problem however is that the sample names which i want to match are a little bit different from the keys in a dictionairy i have from which i need to get the correct value.

example:

sample = "foo_foo.bar.12"
matching_dict = {"foo_foo-bar-12" : "foo.bar.12"}

I have about 5500 samples, each with a different type of arrangement, so not every sample looks like the example i gave. Ideally i want a dynamic way of comparing the 2 strings with eachother and then get the value from the dict if they are most alike.

Thank in advance!

https://stackoverflow.com/questions/17388213/find-the-similarity-metric-between-two-strings probably has some useful info for you. — Tom Dalton, Nov 27 '20 at 15:54

score 2 · Accepted Answer · answered Nov 27 '20 at 15:55

You could use Levenshtein distance. This measures how similar two strings are to eachother. There is a very easy python libarary for it called python-levenshtein. With this you could compare your sample to all the values in the dictionary, and calculate which value in the dict has the lowest Levenshtein distance.

score 0 · Answer 2 · answered Nov 27 '20 at 16:12

As peter wood suggested, you can try fuzzywuzzy. It uses Levenshtein Distance to calculate the differences between sequences in a simple-to-use package. https://github.com/seatgeek/fuzzywuzzy

pip install fuzzywuzzy
from fuzzywuzzy import fuzz
fuzz.ratio("this is a test", "this is a test!")
>>> 97

Is there a way to compare two strings for likeness?

2 Answers2