March 28, 2012 at 9:51 pm
Comments posted to this topic are about the item Find Customers Who Bought "A" and "B" But Not "C" (SQL Spackle)
--Jeff Moden
Change is inevitable... Change for the better is not.
March 28, 2012 at 9:56 pm
Did you compare the speed of this query against one where your first (Group By) From clause is a Select Distinct (sub) query?
Nice article. It builds nicely for the reader.
I love seeing the Except clause being introduced to the reader who, like me, cut our teeth on SQL92 and need to learn newer language features to be more productive.
March 28, 2012 at 10:17 pm
Another way to get the result would be to use Intersect and then combine it with Except. The distinct part is handled implicitly.
--===== Find Customers that bought both "A" AND "B"
SELECT CustomerID
FROM #Purchase
WHERE ProductCode IN ('A')
INTERSECT
SELECT CustomerID
FROM #Purchase
WHERE ProductCode IN ('B')
EXCEPT
--===== Find Customers that bought "C".
SELECT CustomerID
FROM #Purchase
WHERE ProductCode IN ('C')
;
March 28, 2012 at 10:53 pm
Wouldnt something like this work just as well?
SELECT
CustomerId
FROM #Purchase
WHERE ProductCode IN ('A','B', 'C')
GROUP BY CustomerID
having sum(case when ProductCode = 'A' then 1 else 0 end) > 0
and sum(case when ProductCode = 'B' then 1 else 0 end) > 0
and sum(case when ProductCode = 'C' then 1 else 0 end) = 0
/T
March 28, 2012 at 11:04 pm
tommyh (3/28/2012)
Wouldnt something like this work just as well?
SELECT
CustomerId
FROM #Purchase
WHERE ProductCode IN ('A','B', 'C')
GROUP BY CustomerID
having sum(case when ProductCode = 'A' then 1 else 0 end) > 0
and sum(case when ProductCode = 'B' then 1 else 0 end) > 0
and sum(case when ProductCode = 'C' then 1 else 0 end) = 0
/T
Yep, this is how i once implemented it! Cant remember the thread now.
Here it is: http://www.sqlservercentral.com/Forums/FindPost1267224.aspx
Proved to elimate the EXCEPT part; also does a single scan on the table.
March 28, 2012 at 11:53 pm
Nice spackle Jeff, and great alternatives in the discussion!
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
March 28, 2012 at 11:55 pm
OH yeah, Nice one Jeff. As always! Thanks a lot for taking time to teach us some of the tricks of the game. Thaks a lot 🙂
March 29, 2012 at 12:37 am
Had a play with this and was very impressed - nice job.
I played around and added this index
CREATE INDEX IX_#Purchase_ProductCode
ON #Purchase (
ProductCode)
INCLUDE (CustomerID)
One concern I had was looking at the query plan, the EXCEPT does a clustered index scan. That may be a function of the small data set, but I am not sure.
My preference for these types of queries has always been to do an OUTER JOIN on the one we don't want selected and then say that we only want the ones that didn't match returned.
SELECT a.CustomerID
FROM #Purchase a
LEFT OUTER JOIN #Purchase b
ON b.CustomerID = a.CustomerID
AND b.ProductCode IN ('C')
WHERE a.ProductCode IN ('A','B')
AND b.CustomerID IS NULL
GROUP BY a.CustomerID
HAVING COUNT(DISTINCT a.ProductCode) = 2
Any thoughts?
March 29, 2012 at 1:07 am
"Left Join Where Is Null" gets complicated to read/understand and is subject to cardinalities issues that will impact performance if the right set has many records for 1 record in the left hand side.
There is also this way, probably the closest to natural language.
SelectDistinct
CustomerID
From#Purchase
WhereProductCode = 'A'
AndCustomerID In
(
SelectCustomerID
From#Purchase
WhereProductCode = 'B'
)
AndCustomerID Not In
(
SelectCustomerID
From#Purchase
WhereProductCode = 'C'
)
I would be interested to have this table filled by 100,000+ rows of sample data and then compare execution plans and statistics of the proposed approaches.
Comparing syntax without an idea of the performance signature is a bit useless IMHO.
So here is another version of the initialization script from which the difference will be more blatant:
--===== Conditionally drop the test table to make
-- reruns in SSMS easier.
IF OBJECT_ID('tempdb..#Purchase','U') IS NOT NULL
DROP TABLE #Purchase
;
--===== Create the test table
CREATE TABLE #Purchase
(
PurchaseID INT IDENTITY(1,1),
CustomerID INT,
ProductCode VARCHAR(1000)
PRIMARY KEY CLUSTERED (PurchaseID)
)
;
--===== Populate the test table with known data.
INSERT INTO #Purchase
(CustomerID, ProductCode)
------- Customer #1 precisely meets the criteria.
-- Bought 'A' and 'B' but not 'C'.
SELECT 1, 'A' UNION ALL
SELECT 1, 'B' UNION ALL
------- Customer #2 also meets the criteria.
-- Bought 'A' and 'B' and somthing else,
-- but not 'C'.
SELECT 2, 'A' UNION ALL
SELECT 2, 'B' UNION ALL
SELECT 2, 'D' UNION ALL
------- Customer #3 also meets the criteria.
-- Bought 'A' and 'B' and something else,
-- but not 'C'.
SELECT 3, 'A' UNION ALL
SELECT 3, 'B' UNION ALL
SELECT 3, 'D' UNION ALL
SELECT 3, 'A' UNION ALL
SELECT 3, 'D' UNION ALL
------- Customer #4 doesn't meet the criteria.
-- Bought 'A' and 'B' but also bought 'C'.
SELECT 4, 'A' UNION ALL
SELECT 4, 'B' UNION ALL
SELECT 4, 'C' UNION ALL
------- Customer #5 doesn't meet the criteria.
-- Bought 'A' and 'B' and something else,
-- but also bought 'C'.
SELECT 5, 'A' UNION ALL
SELECT 5, 'B' UNION ALL
SELECT 5, 'A' UNION ALL
SELECT 5, 'B' UNION ALL
SELECT 5, 'C' UNION ALL
SELECT 5, 'D' UNION ALL
------- Customer #6 doesn't meet the criteria.
-- Bought more than 1 of 'A' and something else
-- but not 'B'.
SELECT 6, 'A' UNION ALL
SELECT 6, 'A' UNION ALL
SELECT 6, 'D' UNION ALL
SELECT 6, 'E' UNION ALL
------- Customer #7 doesn't meet the criteria.
-- Bought more than 1 of 'B' and something else
-- but not 'A'.
SELECT 7, 'B' UNION ALL
SELECT 7, 'B' UNION ALL
SELECT 7, 'D' UNION ALL
SELECT 7, 'E'
go
declare @i int = 10;
while @i < 100000
begin
insert into #Purchase (CustomerID, ProductCode)
select CustomerID + @i, ProductCode + Left(ProductCode, 1)
From #Purchase
set @i = @i * 2
end
Then turn on statistics or open your SQL profiler
SET STATISTICS IO ON
And now you're ready.
Then you will also need an index on ProductCode.
Create Index IX_ProductCode On #Purchase (ProductCode) Include (CustomerID);
March 29, 2012 at 2:23 am
How about this? Although i agree query might get longer when there would be more than 3 items
SELECT DISTINCT CustomerID
FROM #Purchase P
WHERE EXISTS (SELECT 1 FROM #Purchase P1
WHERE P.customerid = P1.customerid
AND P1.productcode = 'A'
)
AND EXISTS (SELECT 1 FROM #Purchase P2
WHERE P.customerid = P2.customerid
AND P2.productcode = 'B'
)
AND NOT EXISTS (SELECT 1 FROM #Purchase P3
WHERE P.customerid = P3.customerid
AND P3.productcode = 'C'
)
March 29, 2012 at 2:27 am
chintan.j.gandhi (3/29/2012)
How about this? Although i agree query might get longer when there would be more than 3 items
SELECT DISTINCT CustomerID
FROM #Purchase P
WHERE EXISTS (SELECT 1 FROM #Purchase P1
WHERE P.customerid = P1.customerid
AND P1.productcode = 'A'
)
AND EXISTS (SELECT 1 FROM #Purchase P2
WHERE P.customerid = P2.customerid
AND P2.productcode = 'B'
)
AND NOT EXISTS (SELECT 1 FROM #Purchase P3
WHERE P.customerid = P3.customerid
AND P3.productcode = 'C'
)
YOu can combine A and B code chunks and leave the NOT EXISTS as it, reducing it tp 2x slower than the others.
March 29, 2012 at 2:43 am
Great article and discussion. Learned some new stuff, got some "food for thought" - nice one, peeps.
March 29, 2012 at 3:15 am
You need to eliminate hash join and Index scan.
With this query you can do that using INTERSECT AND EXCEPT
SELECT CustomerID
FROM #Purchase
WHERE ProductCode = 'A'
INTERSECT
SELECT CustomerID
FROM #Purchase
WHERE ProductCode = 'B'
EXCEPT
SELECT CustomerID
FROM #Purchase
WHERE ProductCode = 'C'
Or, this one using CTE AND JOINS:
;WITH CTE AS
(
SELECT CustomerID
FROM #Purchase
WHERE ProductCode = 'A'
GROUP BY CustomerID
)
,CTE2 AS
(
SELECT CustomerID
FROM #Purchase
WHERE ProductCode = 'B'
GROUP BY CustomerID
)
,CTE3 AS
(
SELECT CustomerID
FROM #Purchase
WHERE ProductCode = 'C'
GROUP BY CustomerID
)
select CTE.CustomerID from CTE
left join CTE3 ON
CTE.CustomerID = CTE3.CustomerID
where CTE.CustomerID in (SELECT CustomerID FROM CTE2)
AND CTE3.CustomerID IS NULL
GROUP BY CTE.CustomerID
March 29, 2012 at 3:20 am
Dalibor Margotic (3/29/2012)
You need to eliminate hash join and Index scan.With this query you can do that using INTERSECT AND EXCEPT
...
Or, this one using CTE AND JOINS:
...
In the CTE and JOINS query, I would replace the IN clause with an EXISTS clause.
But that could be personal preference.
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
March 29, 2012 at 3:39 am
Another great SQL Spackle, Jeff :w00t:
Koen Verbeeck (3/29/2012)
Dalibor Margotic (3/29/2012)
You need to eliminate hash join and Index scan.With this query you can do that using INTERSECT AND EXCEPT
...
Or, this one using CTE AND JOINS:
...
In the CTE and JOINS query, I would replace the IN clause with an EXISTS clause.
But that could be personal preference.
One should certainly replace the IN by EXISTS or a left join adding a where clause containing the "outer-joined object key " is null
or at least compare their performance 😉
interesting series to read on this matter:
- http://sqlinthewild.co.za/index.php/2010/03/23/left-outer-join-vs-not-exists/
Johan
Learn to play, play to learn !
Dont drive faster than your guardian angel can fly ...
but keeping both feet on the ground wont get you anywhere :w00t:
- How to post Performance Problems
- How to post data/code to get the best help[/url]
- How to prevent a sore throat after hours of presenting ppt
press F1 for solution, press shift+F1 for urgent solution 😀
Need a bit of Powershell? How about this
Who am I ? Sometimes this is me but most of the time this is me
Viewing 15 posts - 1 through 15 (of 166 total)
You must be logged in to reply to this topic. Login to reply