aggregations.py

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

'''
SELECT product_id,
       SUM(num_pieces_sold) AS total_pieces_sold,
       AVG(num_pieces_sold) AS average_pieces_sold,
       MAX(num_pieces_sold) AS max_pieces_sold_of_product_in_orders,
       MIN(num_pieces_sold) AS min_pieces_sold_of_product_in_orders,
       COUNT(num_pieces_sold) AS num_times_product_sold
FROM sales_table
GROUP BY product_id
'''
sales_table_execution_plan = sales_table.groupBy(
    col("product_id")
).agg(
    sum("num_pieces_sold").alias("total_pieces_sold"),
    avg("num_pieces_sold").alias("average_pieces_sold"),
    max("num_pieces_sold").alias("max_pieces_sold_of_product_in_orders"),
    min("num_pieces_sold").alias("min_pieces_sold_of_product_in_orders"),
    count("num_pieces_sold").alias("num_times_product_sold")
)

sales_table_execution_plan.show()