Quick SQL question . . .

Perhaps a little on the hacky side but....

select count(1) from some_table group by identno order by 1 desc limit 1;

If the result is anything other than 1, it's not unique.

Sorry for the hacky nature, this is the "thought about it for 10
seconds" version. :-)

-Fran

Peter E. Chen wrote:

Try

select identno, count(identno) from some_table group by identno having
count(identno) >=1;

-----Original Message-----
From: pgsql-general-owner@postgresql.org
[mailto:pgsql-general-owner@postgresql.org]On Behalf Of Peter E. Chen
Sent: Thursday, May 09, 2002 2:39 PM
To: pgsql-general@postgresql.org
Subject: [GENERAL] Quick SQL question . . .

Can anyone tell me what is the easiest way for me to tell if a column is
unique or not? I tried using DISTINCT ON and COUNT together in a SELECT
statement, but I can't seem to get the query to work:

SELECT DISTINCT ON (identno) count(identno) FROM some_table;

I was trying to figure out if the # of unique entries for a particular
column is equal to the # of total entries for that column.

Any suggestions?

Peter

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster

On Thu, 9 May 2002, Peter E. Chen wrote:

Can anyone tell me what is the easiest way for me to tell if a column is
unique or not? I tried using DISTINCT ON and COUNT together in a SELECT
statement, but I can't seem to get the query to work:

SELECT DISTINCT ON (identno) count(identno) FROM some_table;

I was trying to figure out if the # of unique entries for a particular
column is equal to the # of total entries for that column.

Any suggestions?

SELECT identno, count(*) FROM some_table
GROUP BY 1 HAVING COUNT(*) = 1;

Regards Herbie
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Herbert Liechti http://www.thinx.ch
ThinX networked business services Adlergasse 5, CH-4500 Solothurn
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Is it just me or do the suggestion made look wrong to anyone else?

Anyway, this is the closest to my mind and I think what I think is the mistake
is just a typo. So I would use.

SELECT identno, count(identno)
FROM some_table
GROUP BY identno
HAVING count(identno) > 1

or even

SELECT count(1) FROM (
SELECT count(identno)
FROM some_table
GROUP BY identno
HAVING count(identno) > 1) a

either one of which will return one of more rows if Peter's uniqueness test
fails.

Right, now someone can correct me :)

On Thu, 9 May 2002, Marie G. Tuite wrote:

Try

select identno, count(identno) from some_table group by identno having
count(identno) >=1;

-----Original Message-----
From: pgsql-general-owner@postgresql.org
[mailto:pgsql-general-owner@postgresql.org]On Behalf Of Peter E. Chen
Sent: Thursday, May 09, 2002 2:39 PM
To: pgsql-general@postgresql.org
Subject: [GENERAL] Quick SQL question . . .

Can anyone tell me what is the easiest way for me to tell if a column is
unique or not? I tried using DISTINCT ON and COUNT together in a SELECT
statement, but I can't seem to get the query to work:

SELECT DISTINCT ON (identno) count(identno) FROM some_table;

I was trying to figure out if the # of unique entries for a particular
column is equal to the # of total entries for that column.

Any suggestions?

Peter

--
Nigel J. Andrews
Director

---
Logictree Systems Limited
Computer Consultants