Go Premium for a chance to win a PS4. Enter to Win

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 276
  • Last Modified:

finding max time in calculations

I need to find the max time difference between two times.
I am getting an aggregate error.
SELECT CLE_IDX_CHAN, COUNT(CLE_IDX_OFFHOOK) ''Call Count'', SUM(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)) ''Seconds'',
            CONVERT(VARCHAR(12),DATEADD(ss, AVG(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)), CAST(0 AS DATETIME)),8) ''Avg Length'',
            CONVERT(VARCHAR(12),DATEADD(ss, SUM(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)), CAST(0 AS DATETIME)),8) ''Total Length'',
            MAX(CONVERT(VARCHAR(12),DATEADD(ss, SUM(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)), CAST(0 AS DATETIME)),8) )
            FROM CL_EVENTS_IDX WITH (NOLOCK)
                  INNER JOIN CL_EVENTS A WITH (NOLOCK) ON CLE_IDX_KEY = CLE_ID
                  INNER JOIN UTILITY_INFO WITH (NOLOCK) ON CLE_IDX_UCODE = UTILITY_KEY
                  INNER JOIN CL_EVENTS B WITH (NOLOCK) ON A.CLE_ID = B.CLE_ID AND A.CLE_CALLID = B.CLE_CALLID
                  INNER JOIN CL_APPID_XREF WITH (NOLOCK) on A.CLE_EVID = CLX_APPID
            WHERE A.CLE_TYPE = 508 AND B.CLE_TYPE = 536 and a.CLE_EVID = b.CLE_EVID
                  AND (A.CLE_EVID = 1 OR  A.CLE_EVID = 13)
            Group BY CLE_IDX_CHAN

Open in new window

0
portlight
Asked:
portlight
1 Solution
 
HugoHiaslCommented:
The two lines beginning with CONVERT are not valid aggregations for a group by statement:

CONVERT(VARCHAR(12),DATEADD(ss, AVG(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)), CAST(0 AS DATETIME)),8) ''Avg Length'',
0
 
portlightAuthor Commented:
Those are not giving me an issue. It is when I add the

 MAX(CONVERT(VARCHAR(12),DATEADD(ss, SUM(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)), CAST(0 AS DATETIME)),8) )

Open in new window


Line that I begin to get the problem.
0
 
David ToddSenior DBACommented:
Hi ,

Try this:
select
	cs.CLE_IDX_CHAN
	, cs.callcount as 'Call Count'
	, cs.seconds as seconds
	, convert( varchar( 12 ), dateadd( ss, cs.AvgLength, 0 ), 8 ) as 'Average Length'
	, convert( varchar( 12 ), dateadd( ss, cs.seconds, 0 ), 8 ) as 'Total Length'
	, convert( varchar( 12 ), dateadd( ss, cs.MaxLength, 0 ), 8 ) as 'Max Length'
from (
	SELECT 
		CLE_IDX_CHAN
		, COUNT(CLE_IDX_OFFHOOK) 'CallCount'
		, SUM(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)) 'Seconds'
		, AVG(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)) 'AvgLength'
		, max( datefidd( ss, cle_idx_offhook, cle_idx_onhook )) 'MaxLength'
	FROM CL_EVENTS_IDX WITH (NOLOCK)
	INNER JOIN CL_EVENTS A WITH (NOLOCK) 
		ON CLE_IDX_KEY = CLE_ID
	INNER JOIN UTILITY_INFO WITH (NOLOCK) 
		ON CLE_IDX_UCODE = UTILITY_KEY
	INNER JOIN CL_EVENTS B WITH (NOLOCK) 
		ON A.CLE_ID = B.CLE_ID 
		AND A.CLE_CALLID = B.CLE_CALLID
		and a.CLE_EVID = b.CLE_EVID
	INNER JOIN CL_APPID_XREF WITH (NOLOCK) 
		on A.CLE_EVID = CLX_APPID
	WHERE 
		A.CLE_TYPE = 508 
		AND B.CLE_TYPE = 536 
		AND (A.CLE_EVID = 1 OR  A.CLE_EVID = 13)
	Group BY 
		CLE_IDX_CHAN 
	) cs
;

Open in new window


What I've done:
Moved a line from the where clause to the join of cl_events
Moved the converts to varchar etc to an outer query
Left an inner query with the group by

HTH
  David

PS If you'd like it tested, please provide test data
0
 
Eugene ZCommented:
<I am getting an aggregate error.>
please post the error
0
 
Scott PletcherSenior DBACommented:
>> Line that I begin to get the problem.
MAX(CONVERT(VARCHAR(12),DATEADD(ss, SUM(DATEDIFF(ss, CLE_IDX_OFFHOOK, CLE_IDX_ONHOOK)), CAST(0 AS DATETIME)),8) ) <<

You can't do a MAX() of a SUM() in SQL Server; you can't embed any of the aggregate functions within each other.
0

Featured Post

NEW Veeam Backup for Microsoft Office 365 1.5

With Office 365, it’s your data and your responsibility to protect it. NEW Veeam Backup for Microsoft Office 365 eliminates the risk of losing access to your Office 365 data.

Tackle projects and never again get stuck behind a technical roadblock.
Join Now