optimizing pyspark performance with schema

join shbcf.ru