Repository for a metrical project within the fall-semester "Pandora" seminary series on machine learning and Ancient Greek at Uppsala university.1
I've put together a number of small tools related to the Levenshtein distance between metrical lines in the corpus. Trivially, levenshtein_exe.py can be used to just find the metric lines most Levenshtein-similar to a given line. There are innumerable usecases for this metric, e.g. tracking rhythmic intertextuality, confirming rhythmic genre conventions etc. If you chmod +x and run e.g.
levenshtein_exe.py "u-u-u-u-u-u-"
it will output the ten Levenshtein-closest lines in the corpus.
Footnotes
-
In the below I will take for granted the file structure of the
masterbranch of the present repository, i.e. all mentions of scripts will be relative this folder structure. Specifically, all scripts will tacitly draw upon the corpus of txt-files in the folder hypotactic_txts_greek if unless otherwise stated. ↩