SQL query!

(Please ignore my previous thread.)
I have considered all the requirements below. Please help in formulating an SQL for it. Thanks!

I have a Table with the columns: Column-a, Column-b, Column-c,Column-d.....Column-z.
But, for the below problem, we can ignore the values in the column-d, .....,Column-z.

I would like to form an SQL selecting the distinct values of the Column-a where

• Column-c is NOT NULL
• There is more than 1 distinct association between Column-a, Column-b values, and Column-b values have 1:1 association with the Column-c values in that rows.

E.g.
Column-a Column-b Column-c
-------------- --------- --------

1      10        S1
1      10        S2
2      20        S3
2      30        S4
3      40        S5
4      50       (NULL)

Here,

• '1' should not be selected, as there is only one association exists with Column-b: 1,10
• '2' should be selected, as there are multiple associations exists i.e. 2,20 & 2,30 and 20, 30 have distinct values in Column-c.
• '3' should not be selected, as there is only association exists with Column-b
• '4' should not be selected, as there is a NULL in column 'c'

How do we achieve it in the SQL? Received on Wed Jan 17 2007 - 02:46:41 CST