This seemed necessary because prep_dataframe was being reused in nference.py. This brought a lot of imports with it, and our inference chain failed on missing CatBoost library in OpenEO environment. It seems unnecessary to do all these imports just for processing the parquet files.
This seemed necessary because
prep_dataframewas being reused innference.py. This brought a lot of imports with it, and our inference chain failed on missing CatBoost library in OpenEO environment. It seems unnecessary to do all these imports just for processing the parquet files.