jamesturk / jellyfish

🪼 a python library for doing approximate and phonetic matching of strings.
https://jamesturk.github.io/jellyfish/
MIT License
2.04k stars 157 forks source link

Jaro-Winkler similarity calculation changed between 0.8.8 and 0.8.9 #150

Closed leonright closed 2 years ago

leonright commented 2 years ago

jellyfish.jaro_winkler_similarity() returns different numbers in 0.8.9 compared to 0.8.8 on Python 3.9.5 with C distribution installed.

Verified on both on Windows 10 and on CentOS 7.

# 0.8.8
>>> import jellyfish
>>> jellyfish.jaro_winkler_similarity("123X", "123Y")
<<< 0.8333333333333334

# 0.8.9
>>> import jellyfish
>>> jellyfish.jaro_winkler_similarity("123X", "123Y")
<<< 0.8833333333333334

Could you please look into the issue or let us know the reason for this change in results received?

Thank you

jamesturk commented 2 years ago

see #147