Comparing figures based on relative date

Question

Post reply

Comparing figures based on relative date

Viewing 5 posts - 16 through 19 (of 19 total)

You must be logged in to reply to this topic. Login to reply

Paul Munter Mr or Mrs. 500 Points: 581 More actions · Answer 1

Agree in this scenario it's for a BI application and I doubt it can handle self joins, throw in the fact it's been dimensionally modelled and it adds it a fair bit more complexity. I literally took it back to bare bones in order to try and break it down a bit then build it back up.

I'm too spoiled now with ETL tools and not close enough to the real action anymore 😀

Again thanks for the responses much appreciated.

ScottPletcher SSC Guru Points: 100262 More actions · Answer 2

ScottPletcher

SSC Guru

Points: 100262

February 11, 2013 at 1:12 pm

#1586727

For best performance,

Cluster the table by

LAST_DAY_OF_MONTH

and not the dopey identity column. If you still need the identity (??), you can leave it as the nonclustered pk.

SQL DBA,SQL Server MVP(07, 08, 09) "Money can't buy you happiness." Maybe so, but it can make your unhappiness a LOT more comfortable!

Dwain Camps SSC Guru Points: 86913 More actions · Answer 3

ChrisM@Work (2/11/2013)

Something like this?

DROP TABLE #mytable
CREATE TABLE #mytable (P_KEY INT IDENTITY(1,1) PRIMARY KEY CLUSTERED, LAST_DAY_OF_MONTH DATETIME, Value int, PreviousYearValue INT) INSERT INTO #mytable VALUES('20130131',20,NULL),('20120131',5,NULL),('20130331',40,NULL),('20120331',2,NULL) SELECT ty.*, ly.* FROM #mytable ty LEFT JOIN #mytable ly
ON ly.LAST_DAY_OF_MONTH = DATEADD(yy,-1,ty.LAST_DAY_OF_MONTH)

This may be an alternative:



CREATE TABLE #mytable

    (P_KEY INT IDENTITY(1,1) PRIMARY KEY CLUSTERED,

      LAST_DAY_OF_MONTH DATETIME,

      Value int,

      PreviousYearValue INT)

INSERT INTO #mytable

	VALUES('20130131',20,NULL),('20120131',5,NULL),('20130331',40,NULL),('20120331',2,NULL)

SELECT P_Key=MIN(P_Key), LAST_DAY_OF_MONTH, Value=SUM(Value), PreviousYearValue=SUM(PreviousYearValue)

FROM (

    SELECT P_Key, LAST_DAY_OF_MONTH, Value, PreviousYearValue

    FROM #mytable

    UNION ALL

    SELECT P_Key, DATEADD(year, 1, LAST_DAY_OF_MONTH), NULL, Value

    FROM #mytable) a

GROUP BY LAST_DAY_OF_MONTH

HAVING SUM(Value) IS NOT NULL

DROP TABLE #mytable

My mantra: No loops! No CURSORs! No RBAR! Hoo-uh![/I]

My thought question: Have you ever been told that your query runs too fast?

My advice:
INDEXing a poor-performing query is like putting sugar on cat food. Yeah, it probably tastes better but are you sure you want to eat it?
The path of least resistance can be a slippery slope. Take care that fixing your fixes of fixes doesn't snowball and end up costing you more than fixing the root cause would have in the first place.

Need to UNPIVOT? Why not CROSS APPLY VALUES instead?[/url]
Since random numbers are too important to be left to chance, let's generate some![/url]
Learn to understand recursive CTEs by example.[/url]
[url url=http://www.sqlservercentral.com/articles/St

ChrisM@Work SSC Guru Points: 186127 More actions · Answer 4

ScottPletcher (2/11/2013)

For best performance,

Cluster the table by

LAST_DAY_OF_MONTH

and not the dopey identity column. If you still need the identity (??), you can leave it as the nonclustered pk.

+1

^{“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

MMartin1 One Orange Chip Points: 27655 More actions · Answer 5

Agree in this scenario it's for a BI application and I doubt it can handle self joins

Maybe I am reading you wrong but I take it you are writing the logic in the application and performance is your concern? Have the application call a sql server stored procedure where the user can supply the date range for the normal dates (not the ones one year ago).

----------------------------------------------------