Question

So I have a table with three columns – col1, col2 and col3. I need to select those values of col1 that share rows with only one combinarion of col2 and col3. In ideal world I would write this:

SELECT col1 FROM table
GROUP BY col1
HAVING COUNT(DISTINCT(col2, col3)) = 1

How to do that in the real world?

So far I have two solutions – group by col1 and col2 and by col1 and col3 them join results; or group by all three then group that by col1. Unfortunately I get (very) different wor counts.

The database is on SQL Server 2005.

Answer 1

SELECT col1
FROM table
GROUP BY col1
HAVING MAX(col2) = MIN(col2) AND MAX(col3) = MIN(col3)

More complex:

SELECT col1
FROM
    (
    SELECT
        col1, 
        ROW_NUMBER() OVER (PARTITION BY col1 ORDER BY col2, col3) as rn
    FROM table
    ) T
GROUP BY col1
HAVING MAX(rn) = 1