SQL & PL/SQL

Hi

I'm going to do some testing, and for that I require to retrieve some data based on a single column e.g test_data_col, which -
1. Has 3 or more count(test_data_col) for a given set of group by columns e.g grp_col1, grp_col2, grp_col3
2. Within the set of rows retrieved, that particular column holds some duplicate values. I don't need the duplicates displayed, just know if duplicates exist or not.

This might explain what I'm trying to do -

grp_col1, grp_col2, grp_col3, test_data_col

1, A, xyz, HELLO
1, A, xyz, HELLO
1, A, xyz, BYE
1, A, xyz, GOODBYE

2, C, pqr, WELCOME
2, C, pqr, GOOD MORNING
2, C, pqr, BAD MORNING

So for condition 1, I do something like this -

SELECT COUNT(test_data_col) cnt, grp_col_1, grp_col2, grp_col3
FROM test_tab
GROUP BY grp_col_1, grp_col2, grp_col3
HAVING COUNT(test_data_col) >= 3;

In this same query, I want to do something that will tell me if the aggregate COUNT(test_data_col) has any duplicate values within it. Again, displaying the duplicates is not important here.

SELECT COUNT(test_data_col) cnt, grp_col_1, grp_col2, grp_col3,
/*some logic*/ dup_val
FROM test_tab
GROUP BY grp_col_1, grp_col2, grp_col3
HAVING COUNT(test_data_col) >= 3;

With the proper coding to replace /*some logic*/, I get following values -

cnt, grp_col_1, grp_col2, grp_col3, dup_val

4, 1, A, xyz, Y
3, 2, C, pqr, N

I just gave dup_val column to explain what I'm trying to achieve.. any other way to know the existence of duplicates in the count aggregate will be fine.

My Oracle version is Oracle Database 11g Enterprise Edition Release 11.1.0.7.0

Appreciate you response!