I have a function that needs to take in a PySpark dataframe with some unknown schema and group it by the 1 column guaranteed to exist ("dt") and write each subset
