joins.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 a.*,
       b.*
FROM sales_table a
    LEFT JOIN sellers_table b
        ON a.seller_id = b.seller_id
'''
#   Left join
left_join_execution_plan = sales_table.join(sellers_table, 
                   on=sales_table["seller_id"] == sellers_table["seller_id"], 
                   how="left")

#   Inner join
inner_join_execution_plan = sales_table.join(sellers_table, 
                   on=sales_table["seller_id"] == sellers_table["seller_id"], 
                   how="inner")

#   Right join
right_join_execution_plan = sales_table.join(sellers_table, 
                   on=sales_table["seller_id"] == sellers_table["seller_id"], 
                   how="right")

#   Full Outer join
full_outer_join_execution_plan = sales_table.join(sellers_table, 
                   on=sales_table["seller_id"] == sellers_table["seller_id"], 
                   how="full_outer")