Question

I want to perform a simple sql query but cannot manage to get it right.

(Sorry for the example, that’s all I could think of to illustrate my problem)

Let’s say I have a table FooBar that contains X columns, but only 3 columns that are useful to me.

DogKind_Id, FoodQuantity, Day_Id

And I want to know, for each DogKind_Id, what was the day where the FoodQuantity was the maximum !
So for a FooBar table like :

DogKind_Id, FoodQuantity, Day_Id 
1,2, 2
1,10,1
1,4,5
2,3, 3
2,9,1
2,87,5
3,0, 3
3,0,2
3,1,6

The result of my query will be :

DogKind_Id, Day_Id
1,1
2,5
3,6

So far, I succeeded in doing this :

SELECT DogKind_Id, MAX(FoodQuantity)
FROM FooBar
GROUP BY DogKind_Id

But I cannot manage to simply use this result to get the Day_Id… If I had Day_Id in the above SELECT, it adds up lots of rows… I tried SUBSELECT, EXISTS but obviously doing it wrong !

Can you guys help me a bit ?

Answer 1

SELECT parent.DogKind_id, parent.Day_Id
FROM FooBar AS parent
INNER JOIN (
    SELECT DogKind_Id, MAX(FoodQuantity) AS Max
    FROM FooBar
    GROUP BY DogKind_Id
) AS child ON (parent.DogKind_Id=child.DogKind_Id) AND (parent.FoodQuantity = child.Max)

basically, the inner query is your version, which gets the id/max values.
You then have the outer query joining against this subquery to figure out WHICH of the records has that max foodquantity, so you can get the associated dayid