"Sam Mason" <sam@samason.me.uk> writes:
> SELECT i, MIN(k) OVER (PARTITION BY j)
> FROM tbl
> GROUP BY i;
>
> This is obviously wrong, but I don't see how to get to where I need to
> be.
I'm not entirely sure myself. I think it might involve RANK OVER j though.
I suspect it will look more like the DISTINCT ON solution than the min(record)
solution.
-- Gregory Stark EnterpriseDB http://www.enterprisedb.com Ask me about EnterpriseDB's RemoteDBA services!