Search code examples
sqlpostgresqlgroup-byhaving

SQL Query SUM with conditions


I'm having some difficulties finding the good query I need, looking for help. After multiple JOIN and a lot of WHERE, I have this columns and values :

+---------+----------+--------+-------+
| id_test | quantity | price  |  fee  |
+---------+----------+--------+-------+
| "1-3"   |        1 | 33.52  | 29.00 |
| "1-5"   |        1 | 33.52  | 29.00 |
| "1-6"   |        1 | 33.52  | 29.00 |
| "1-8"   |        1 | 86.49  | 29.00 |
| "19-1"  |        1 | 176.54 | 29.00 |
| "19-4"  |        1 | 176.54 | 29.00 |
| "19-5"  |        4 | 176.54 | 29.00 |
| "19-6"  |        1 | 199.47 | 29.00 |
| "19-6"  |        1 | 176.54 | 29.00 |
| "20-10" |        2 | 72.67  | 29.00 |
| "20-11" |        2 | 18.95  | 29.00 |
| "20-9"  |        1 | 22.13  | 29.00 |
+---------+----------+--------+-------+

Each id_test object as a quantity, a price and a fee. I would like to SUM everything so to have a global price : (quantity*price)+fee The thing is, and that's where I'm stuck, the fee has to be added only once by id_test. Yet, I have here two id_test "19-6".

If I SUM everything I have 1827.67 (price*quantity) + 348 (fee).

Since I have two "19-6", I need 1827.67 (price*quantity) + 319 (fee).

How would you do in SQL ?

Thanks


Solution

  • You can make use of Common Table Expressions (CTE)

    Assuming your data in "data" table:

    with total_fee as (
        select distinct id_test, fee from data
    ),
    total_price as (
        select distinct id_test, sum(quantity) quantity, sum(price) price 
        from data
        group by id_test
    )
    
    
    select sum(b.price*quantity) quantity, sum(a.fee) fee
    from total_fee a
    left join total_price b
    on a.id_test = b.id_test