Click here to monitor SSC
SQLServerCentral is supported by Red Gate Software Ltd.
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase ««12

SUM CASE, COLUMN MINUS COLUMN Expand / Collapse
Author
Message
Posted Thursday, May 2, 2013 9:37 AM


Ten Centuries

Ten CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen Centuries

Group: General Forum Members
Last Login: Friday, August 22, 2014 8:05 AM
Points: 1,222, Visits: 2,545
Sean Lange (5/1/2013)
ChrisM@Work (5/1/2013)
Sean Lange (5/1/2013)
So what exactly is the issue here?

I would suggest you use the newer join constructs. With this query you don't even need a where clause.

Here is the syntax:

from batt_State s
join pnLU as p on s.PartNo = p.PartNo
group by s.partNo, p.descript



I'd change that to:
	SELECT s.partNo, p.descript,
CLNT = SUM(CASE WHEN battState = '99' THEN qty ELSE 0 END),
AVAIL = SUM(CASE WHEN battState = '1' THEN qty ELSE 0 END),
QAH = SUM(CASE WHEN battState = '20' THEN qty ELSE 0 END),
BOOST = SUM(CASE WHEN battState = '30' THEN qty ELSE 0 END),
RESTING = SUM(CASE WHEN battState = '31' THEN qty ELSE 0 END)

FROM batt_State s
INNER JOIN pnLU p
ON s.PartNo = p.PartNo
WHERE battState IN ('1','20','30','31','99')
GROUP BY s.partNo, p.descript



Oh sure...that will only be better if you want the query to be faster.


That's also assuming that the business requirement is to return rows only where battState IN ('1','20','30','31','99'), which isn't stated by the OP. If he wants a row for every partNo in the batt_State table, but only wants the aggregate SUM values where battState IN ('1','20','30','31','99') (leaving 0 in these columns for other values of battState), then the WHERE battState IN ('1','20','30','31','99') clause should be omitted.


Jason Wolfkill
Blog: SQLSouth
Twitter: @SQLSouth
Post #1448862
Posted Friday, May 3, 2013 2:40 AM


SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Friday, September 12, 2014 9:43 AM
Points: 7,284, Visits: 13,824
wolfkillj (5/2/2013)
Sean Lange (5/1/2013)
ChrisM@Work (5/1/2013)
Sean Lange (5/1/2013)
So what exactly is the issue here?

I would suggest you use the newer join constructs. With this query you don't even need a where clause.

Here is the syntax:

from batt_State s
join pnLU as p on s.PartNo = p.PartNo
group by s.partNo, p.descript



I'd change that to:
	SELECT s.partNo, p.descript,
CLNT = SUM(CASE WHEN battState = '99' THEN qty ELSE 0 END),
AVAIL = SUM(CASE WHEN battState = '1' THEN qty ELSE 0 END),
QAH = SUM(CASE WHEN battState = '20' THEN qty ELSE 0 END),
BOOST = SUM(CASE WHEN battState = '30' THEN qty ELSE 0 END),
RESTING = SUM(CASE WHEN battState = '31' THEN qty ELSE 0 END)

FROM batt_State s
INNER JOIN pnLU p
ON s.PartNo = p.PartNo
WHERE battState IN ('1','20','30','31','99')
GROUP BY s.partNo, p.descript



Oh sure...that will only be better if you want the query to be faster.


That's also assuming that the business requirement is to return rows only where battState IN ('1','20','30','31','99'), which isn't stated by the OP. If he wants a row for every partNo in the batt_State table, but only wants the aggregate SUM values where battState IN ('1','20','30','31','99') (leaving 0 in these columns for other values of battState), then the WHERE battState IN ('1','20','30','31','99') clause should be omitted.


You're absolutely right, Wolfie - but as an experienced dev, I'd bet that you'd write the same as I did, or possibly both, and offer it to a stakeholder to choose.


“Write the query the simplest way. If through testing it becomes clear that the performance is inadequate, consider alternative query forms.” - Gail Shaw

For fast, accurate and documented assistance in answering your questions, please read this article.
Understanding and using APPLY, (I) and (II) Paul White
Hidden RBAR: Triangular Joins / The "Numbers" or "Tally" Table: What it is and how it replaces a loop Jeff Moden
Exploring Recursive CTEs by Example Dwain Camps
Post #1449078
Posted Friday, May 3, 2013 5:23 AM


Ten Centuries

Ten CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen Centuries

Group: General Forum Members
Last Login: Friday, August 22, 2014 8:05 AM
Points: 1,222, Visits: 2,545
ChrisM@Work (5/3/2013)
wolfkillj (5/2/2013)
Sean Lange (5/1/2013)
ChrisM@Work (5/1/2013)
Sean Lange (5/1/2013)
So what exactly is the issue here?

I would suggest you use the newer join constructs. With this query you don't even need a where clause.

Here is the syntax:

from batt_State s
join pnLU as p on s.PartNo = p.PartNo
group by s.partNo, p.descript



I'd change that to:
	SELECT s.partNo, p.descript,
CLNT = SUM(CASE WHEN battState = '99' THEN qty ELSE 0 END),
AVAIL = SUM(CASE WHEN battState = '1' THEN qty ELSE 0 END),
QAH = SUM(CASE WHEN battState = '20' THEN qty ELSE 0 END),
BOOST = SUM(CASE WHEN battState = '30' THEN qty ELSE 0 END),
RESTING = SUM(CASE WHEN battState = '31' THEN qty ELSE 0 END)

FROM batt_State s
INNER JOIN pnLU p
ON s.PartNo = p.PartNo
WHERE battState IN ('1','20','30','31','99')
GROUP BY s.partNo, p.descript



Oh sure...that will only be better if you want the query to be faster.


That's also assuming that the business requirement is to return rows only where battState IN ('1','20','30','31','99'), which isn't stated by the OP. If he wants a row for every partNo in the batt_State table, but only wants the aggregate SUM values where battState IN ('1','20','30','31','99') (leaving 0 in these columns for other values of battState), then the WHERE battState IN ('1','20','30','31','99') clause should be omitted.


You're absolutely right, Wolfie - but as an experienced dev, I'd bet that you'd write the same as I did, or possibly both, and offer it to a stakeholder to choose.

For something this short, I probably would just write it and show the stakeholder the results with and without the WHERE clause. Anything more complex, I'd want clarification of the requirements first - I don't like to spend too much time coding after phantom requirements.


Jason Wolfkill
Blog: SQLSouth
Twitter: @SQLSouth
Post #1449137
« Prev Topic | Next Topic »

Add to briefcase ««12

Permissions Expand / Collapse