Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations SkipVought on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

SQL SERVER 200 and GROUP BY and HAVING with COUNT 1

Status
Not open for further replies.

davism

MIS
Nov 9, 2002
140
0
0
US
Hi all,

I'm having a little bit of difficult and hopefully somebody can help me out here.

I have a table like:

GUID, TESTID, TESTDATE, TESTTYPE, TESTTYPEID

Now what I want to do is I know there are multiple rows with the same TESTID for the same TESTTYPEID.

So, I am trying to get an output of those for the same TESTID.

I am trying this:

SELECT TESTID, TESTDATE from MyTestTable WHERE TestTypeID=1100 and (TESTDATE >= '2014-09-03' and TESTDATE < '2014-09-04')
GROUP BY TESTID, TESTDATE
HAVING COUNT(TESTID) >=2

Now, this is not returning anything but I know for a fact there are records in the table that meet that situation. If I take out the HAVING clause then it will show them to and I see it but I also see all the others too. I only want the ones that have the COUNT(TESTID) >= 2

What am I doing incorrectly here?

This is SQL SERVER 2000.

Thanks
 
Take TestDate out of the field list and the GROUP BY list.

Tamar
 
Hi,

This what you're after?...

Code:
declare @temp table (guid int, testid int, testtypeid int)
insert into @temp select 1, 1, 1
insert into @temp select 2, 1, 1
insert into @temp select 3, 1, 2
insert into @temp select 5, 1, 3
insert into @temp select 6, 1, 3
insert into @temp select 7, 1, 4

select * 
from @temp t1
join (
	select testid, testtypeid
	from @temp
	group by testid, testtypeid
	having count(0) > 1
) t2 on t1.testid = t2.testid and t1.testtypeid = t2.testtypeid

Ryan
 
Hi RyanEK

No, unfortunately, that is not going to work.

If you take the example of:

Code:
insert into @temp select 2, 1, 1
insert into @temp select 3, 1, 2
insert into @temp select 5, 1, 3
insert into @temp select 6, 1, 3
insert into @temp select 7, 1, 4
insert into @temp select 8, 2, 3
insert into @temp select 9, 3, 5
insert into @temp select 10, 1, 2

I would want the information for like:

3, 1, 2
10, 1, 2

Because I want to list TestID's of 1 with a TestTypeID of 2.

Obviously there would be a TestDate with row as well but for the sake of brevity, the TestGUID would work in this case (i.e. the 3 and the 10, in the example.)

I hope that clears it up some.

But how can I do that?
 
but code below produce
3, 1, 2
10, 1, 2

Code:
declare @temp table (guid int, testid int, testtypeid int)
insert into @temp select 2, 1, 1
insert into @temp select 3, 1, 2
insert into @temp select 5, 1, 3
insert into @temp select 6, 1, 3
insert into @temp select 7, 1, 4
insert into @temp select 8, 2, 3
insert into @temp select 9, 3, 5
insert into @temp select 10, 1, 2 

select t1.* 
from @temp t1
join (
	select testid, testtypeid
	from @temp
	group by testid, testtypeid
	having count(0) > 1
) t2 on t1.testid = t2.testid and t1.testtypeid = t2.testtypeid
 
Correct, but it also produced:

5, 1, 3
6, 1, 3

BUT that was an easy fix as with the "on" condition I added "...and t1.testtypeid=2".

The results sure enough only gave me the

3, 1, 2
10, 1, 2

Let me review this a little more and apply it.
 
That is working out great!

Thanks so much and very much appreciate that.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top