Question

Having this table:

Row    Pos      Outdata    Mismatch          Other
1        10       S           0               A
2        10       S           5               A
3        10       R           0               B
4        10       R           7               B
5        20       24          0               A
6        20       24          5               B
6        20       32          10              C

How can I select all rows where Pos=10 having unique Outdata. If more than one row exists, I would like to have the row where the field Mismatch is smallest. Ie I would like to get rows 1 and 3, not 2 and 4.

In that select I would also like to do the same for all Pos=20, so the total result should be rows 1,3,5,6

(And I want to then access the “Other” field, so I cant only SELECT DISTINCT on Pos and OutData and Mismatch).

Is there a query to do this in MySQL?

Answer 1

Rationale is to create a table with all values of Pos, OutData and the lowest Mismatch and use the combination of these fields as a unique key into your actual table.

SELECT    t1.*
FROM      MyTable t1
          INNER JOIN (
            SELECT    Pos, OutData, Mismatch = MIN(Mismatch)
            FROM      MyTable
            GROUP BY  Pos, OutData
          ) t2 ON t2.Pos = t1.Pos
                  AND t2.OutData = t1.OutData
                  AND t2.Mismatch = t1.Mismatch