Question

I need help with the following SubQuery in MySQL

 SELECT ... # Main query 
...      
... IN # Start of subquery that expects to return an array of a.id 
(SELECT a.id, a.multipler 
FROM a 
JOIN b ON a.id = b.id
GROUP BY(a.id)
HAVING COUNT(b.id) * a.multipler < 5)

What I want to achive is to group all a.id that exist in table b (b.id).
I want to count the results and mulitple with an unique mulitpler (a.multipler) for each a.id.

The problem here is that I want this to be a subquery: thus I cannot have 2 results. Though, in order to use “HAVING” I need to have each variable in the result set.

I only want/need “a.id” without “a.multipler” as the result. Any ideas to solve this?

Example code:

#table a
+------+-------------+
| a_id | a_multipler |  
+------+-------------+
|    1 | 2.000       |
|    2 | 0.560       |
|    3 | 1.000       |
|    4 | 1.200       |
|    5 | 2.000       | 
+----- +-------------+
#table b
+------+
| b_id | 
+------+
|    1 |
|    1 | 
|    1 |
|    4 | 
|    4 |
|    3 | 
+------+

# a.id 1: occurance in "table b": 3 x 2.000 ("a.id"==1 "a.multipler"). Fails, result >= 5

#a.id 2: Fails, no occurance of a.id in "table b".

#a.id 3: 1 x 1.000. Result < 5 OK

# and so on... "id" in "table a" (but not in "table b") is unique, 
# also a "id" in "table b" have to exist in "table a".

Wanted result given the above query (a.id): 4,3
NOT wanted result:

(a.id): 4,3

(a.multipler): 1.200, 1.000

Answer 1

Then, wrap your subquery in another SELECTthat selects just the column you need:

SELECT ... # Main query 
...      
... IN
(SELECT pairs.id FROM               # <===          
    (SELECT a.id, a.multipler 
     FROM a 
     JOIN b ON a.id = b.id
     GROUP BY(a.id)
     HAVING COUNT(b.id) * a.multipler < 5) pairs)

Any (sub)query results in a temporary table that can be further manipulated, in this case simply by selecting a subset of it columns.