DISTINCT within aggregates.

Hi,

I have PG 6.4 which does not appear to support DISTINCT with aggregates.

Here's an example with the following table:

key|numero
---+-----
1|one
2|two
3|three
4|one
4|four
2|one
(6 rows)

brecard10=> select count(numero) from test;
count
-----
6
(1 row)

brecard10=> select count(numero) from test group by value;
count
-----
1
3
1
1
(4 rows)

brecard10=> select count(distinct numero) from test group by value;
ERROR: parser: Syntax error at or near "distinct"

I want to be able to count how many DISTINCT 'numero' entries there are in
test (i.e 4).
The first statement is a simple count.
The second uses a group which does indeed tell us that there are four
different 'numero's but does not return a unique count.
The last statement is the statement to use (at least according to my SQL
handbook - Bowman et al.) but is not supported by PG 6.4.

Has anybody found a workaround?

cheers,

Stuart.

+-------------------------+--------------------------------------+
| Stuart Rison | Ludwig Institute for Cancer Research |
+-------------------------+ 91 Riding House Street |
| Tel. (0171) 878 4041 | London, W1P 8BT, UNITED KINGDOM. |
| Fax. (0171) 878 4040 | stuart(at)ludwig(dot)ucl(dot)ac(dot)uk |
+-------------------------+--------------------------------------+