Parquet is a column-based file format which is used to optimize the speed of queries and is very efficient than a CSV or JSON file format. Spark SQL supports both read and write functions on parquet files which capture schema of original data automatically

BY Best Interview Question ON 06 Feb 2020