With the availability of things like pyspark the grey zone between pandas scale and pyspark scale is small and uncommon though. Especially for the awkward tabul

With the availability of things like pyspark the grey zone between pandas scale and pyspark scale is small and uncommon though. Especially for the awkward tabular data manipulation tasks where you actually need to be mapping custom functions and what not. Pandas built ins can cover everything with good performance except dataframe.apply imo That and using dicts as maps.