search

ClimbsRocks / auto_ml

[UNMAINTAINED] Automated machine learning for analytics & production
http://auto-ml.readthedocs.io
MIT License
1.64k stars 310 forks source link

categorical column with float and np.nan throws error when being converted to strings only with lgbm #315

Open ClimbsRocks opened 7 years ago

ClimbsRocks commented 7 years ago

maybe try explicitly casting to strings any column that should be categorical?

samching commented 7 years ago

Not sure if it's related to the above error, but I'm getting the following error while running LightGBM on the get_boston_dataset() test set.

TypeError: 'NoneType' object is not iterable as well as TypeError: Unknown type(ndarray)

ClimbsRocks commented 7 years ago

@samuelcwl : bah, that was a really, really annoying error. lightgbm is generally run as an awesome engineering project, but in this case, they deprecated one of their formerly accepted data formats, without a deprecation warning.

the second of those two errors should be fixed in the release i just finished and pushed to pypi (v2.6.0). but let me know if you keep encountering that first issue, or any other issues, and i'll try to fix them up asap!

i also love success stories and general use stories, so if you've got any, i'm all ears.