left_anti.py

#   Read the source tables in Parquet format
sales_table = spark.read.parquet("./data/sales_parquet")
sellers_table = spark.read.parquet("./data/sellers_parquet")

'''
SELECT *
FROM sales_table
WHERE seller_id NOT IN (SELECT seller_id FROM sellers_table)
'''
#   Left Anti joins are a way to express the NOT IN operation in SQL
anti_join_execution_plan = sales_table.join(sellers_table,
                on=sales_table["seller_id"] == sellers_table["seller_id"],
                how="left_anti")

anti_join_execution_plan.show()