Optimize lexicon - Githubissues

nalmadi / Namesake

A Python Checker for Lexical Similarity in Identifier Names

MIT License

6 stars 4 forks source link

Optimize lexicon #4

Closed minhphd closed 1 year ago

minhphd commented 1 year ago

Some changes:

created readable lexicon in csv format, along with python file to generate lexicon csv from pickle file
fixed dir path in letter_congruence_multiprocessing.py, fixed issue #1
modified get_congruence() to take into account rotation when finding orthographic similarities between pairs, tested on 5 different pairs

Some notes for future:

regenerate complete lexicon file with new get_congruence() method
find better equation to get the final orthographic similarity y value

minhphd commented 1 year ago

redo pull request, too many changes