left_semi.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 IN (SELECT seller_id FROM sellers_table)
'''
#   Left Semi joins are a way to express the IN operation in SQL
semi_join_execution_plan = sales_table.join(sellers_table, 
                on=sales_table["seller_id"] == sellers_table["seller_id"], 
                how="left_semi")

semi_join_execution_plan.show()