Domanda di colloquio di HCLTech

Why do we use parquet file format ? Why dataframes are faster than RDD