>Suppose you have issued a sql-select which returns several records containing the fields COST and SITE among others. Then suppose you want to select from this result the minimum COST and its corresponding SITE. As in:
>

>select min(cost), site from crs1
>

>
>This produces an error Group clause invalid. How should this second select be structured?
>
>
>Thanks for your suggestions.

select crs1.* from crs1 inner join ;
(select site, MIN(cost) as MinCost from crs1 group by 1) crs2 on ;
crs1.site = crs2.site and crs1.Cost = crs2.MinCost

Note, that this will return several records if you have several identical Min cost.