Re: JOIN and aggregate problem

On Fri, 20 Feb 2009, Tarlika Elisabeth Schmitz wrote:

> I have 2 tables T1 and T2
>
> T1 has the columns: D, S, C. The combination of D,S,C is unique.
> T2 has the columns: D, S, C, and boolean X. The combination of D,S,C is
> not unique.
>
> I need to produce the following result for every occurrence of T1:
> D,S,C, COUNT
>
> COUNT is the number of matching D,S,C combinations in T2 where X = true.
> There might be no matching pair in T2 or there might be match but X
> is false.
>
> How can I express this?

Maybe something like one of these barely tested queries?

select d, s, c, sum(case when t2.x then 1 else 0 end)
from t1 left outer join t2 using(d,s,c)
group by d, s, c;

or

select d,s,c,(select count(*) from t2 where t2.d=t1.d and t2.s=t1.s and t2.c=t1.c and t2.x)
from t1;