Is an ontology “better” than a relational data model? “More expressive power” doesn’t always mean “better”.

Leave beacons in your code. I would have avoided a silly error if a variable named xgb_train_data would have been named, for example, xgb_train_data_filepath instead.

Add a CITATION.cff file to your git repository. The Citation File Format is automatically rendered on GitHub and usable by Zenodo and Zotero.

Datasets are easier to reuse if they use standards that are well-established, particularly in a given domain.