That’s happening because column m.unit is used in the SELECT clause but not in the GROUP BY clause.

This is a perfectly valid and reasonable response from our DBMS, but in this exceptional case, selecting columns without including a non-aggregate column in the group by clause would be handy.

But this can’t happen if we are not explicit in what we ask for :

SELECT
 a.account_category,
 a.account_id,
 sum (b.value) as summed_amount 
FROM (
 SELECT m.material_id, m.year
 CASE 
  WHEN m.unit = 'PIECES'
   THEN m.mean_value * r.quantity * m.ratio 
   ELSE m.mean_value * r.quantity / m.ratio
  END as value
 from materials m, requests r
 where r.year = '1/1/13' and
       r.year = m.year and
       r.material_id = m.material_id) b,
                              accounts a
WHERE
    a.material_id = b.material_id and
    a.year = b.year
GROUP BY account_category, account_id

This works because of an inline view (marked in red) which acts as a temporary table holding the value for each row :

and exporting material_id and year to the outside scope, both of which will subsequently used for joining with the Accounts table of the outer query ( a.material_id = b.material_id and a.year = b.year), producing :

and finally grouped into :

enabling us to answer questions like “how much money on average, did the consumption of fruit cost us this year?”, so that we can estimate how much money we should reserve for next year’s purchases.

Other SQL Workshops

SQL Workshop - Selecting columns without including a
non-aggregate column in the group by clause

SQL Workshop - Subselects And Join

More SQL Workshop soon!