Question

Here is my table data from which i want to find a specific id by comparing multiple rows..

onetsoc_code    element_id   data_value

11-1011.00      1.B.1.a      1.33
11-1011.00      1.B.1.b      2.00
11-1011.00      1.B.1.c      2.67
11-1011.00      1.B.1.d      3.67
11-1011.00      1.B.1.e      7.00
11-1011.00      1.B.1.f      5.33
11-1011.00      1.B.1.g      5.00
11-1011.00      1.B.1.h      6.00
11-1011.00      1.B.1.i      0.00
11-1021.00      1.B.1.a      1.33
11-1021.00      1.B.1.b      1.33
11-1021.00      1.B.1.c      1.00
11-1021.00      1.B.1.d      3.33
11-1021.00      1.B.1.e      7.00
11-1021.00      1.B.1.f      3.67
11-1021.00      1.B.1.g      5.00
11-1021.00      1.B.1.h      6.00
11-1021.00      1.B.1.i      4.00
11-1031.00      1.B.1.a      1.00
11-1031.00      1.B.1.b      3.67
11-1031.00      1.B.1.c      3.67
11-1031.00      1.B.1.d      4.67
11-1031.00      1.B.1.e      7.00

This is my table so i want those onetsoc_code where the

1.B.1.g      5.00
1.B.1.h      6.00
1.B.1.i      4.00

If you look at the table then you will find 11-1021.00 is the result which i want.

So this is the logic what i want you for got 1.B.1.a to 1.B.1.f. I only want to compare three rows at a moment and want the single onetsoc_code associated with those data.

The rows with 1.B.1.g, 1.B.1.h, 1.B.1.i is available for all onetsoc_code so i just need those onetsoc_code which has above given condition like again,

1.B.1.g      5.00
1.B.1.h      6.00
1.B.1.i      4.00

Answer 1

You can group by onetsoc_code and filter for those groups that contain three DISTINCT records:

SELECT   onetsoc_code
FROM     mytable
WHERE    (element_id = '1.B.1.g' AND data_value = 5.00)
      OR (element_id = '1.B.1.h' AND data_value = 6.00)
      OR (element_id = '1.B.1.i' AND data_value = 4.00)
GROUP BY onetsoc_code
HAVING   COUNT(DISTINCT element_id, data_value) = 3

See it on sqlfiddle.

Of course, if uniqueness is guaranteed one can save the expense of the DISTINCT operation and just use plain old COUNT(*) instead.