You are provided with the sales and returns for a big-box office supply store. The SALES DATA is in the form of rowID, order number, order date, order quantity, total sales, customer name, state, region, product, and ship date. The RETURNS DATA is in the form of order number and status. There is a concern about the number of returns being generated. Please provide the following using Hadoop Hive / SQL-type coding format without a specificed data set:

1. The aggregate amount of returned sales (rounded to the nearest dollar), and number of returns by region.

2. The top 5 (only) customers who have returned products based on aggregate sales (rounded to the nearest dollar) as well as the number of sales and the region.

3. The top 5 (only) products that are returned based on aggregate sales (rounded to the nearest dollar) as well as the number of sales.