If I have a dataframe with the following columns:
1. NAME object 2. On_Time object
If after 6 years you still have the issue, this should solve it :)
cols = [c for c in df.columns if df[c].dtype in ['object', 'datetime64[ns]']]