Number of partitions created when I output a parquet file

[saveenrMSFT]

How can I control the number of partitions created when I output a parquet file?

[chetnachaudhari · Accepted solution]

Hi @saveenrMSFT,

  If you are using PySpark, you can control the number of partitions created when you output a Parquet file by using the repartition method or the coalesce method on your DataFrame before writing it to Parquet. These methods allow you to control the number of output partitions, which in turn affects the number of Parquet files generated.

Thanks,

Chetna