Running Totals on PIVOT in T-SQL

  • I have the following query

    SELECT * FROM (

    SELECT Product

    ,[201101],[201102],[201103],[201104],[201105],[201106],[201107],[201108],[201109],[201110],[201111],[201112]

    FROM

    (SELECT Product, [TransactionYYYYMM1]

    FROM ##tmpTable T1 LEFT JOIN Products T2 ON T1.ID = T2.ID WHERE Sold = 1

    ) AS source

    PIVOT (

    COUNT(TransactionYYYYMM1) FOR [TransactionYYYYMM1] IN ([201101],[201102],[201103],[201104],[201105],[201106],[201107],[201108],[201109],[201110],[201111],[201112])) AS pvt

    ) T1

    ORDER BY 1

    which produces the data in the "current" section of the attached speadsheet. What I want is to do totalling within my query (see "desired" section). Any ideas? I have been messing with it for a few days and no luck.

    Thanks,

    Michael

  • GROUP BY with ROLLUPs may be?

  • I have tried that, I didn't have a whole lot of success...but odds are, I am doing it incorrectly.

  • Can you post the query you tried ?

    And also, can you post the data in readily-consumable format? I cannot download content at my office.

    Please read here on how ro format data to be readily consumable - http://www.sqlservercentral.com/articles/Best+Practices/61537/

  • Thanks for the information....lets see if this helps

    IF OBJECT_ID('TempDB..#mytable','U') IS NOT NULL

    DROP TABLE #mytable

    CREATE TABLE #mytable(

    [Product] [varchar](1) NULL,

    [TransactionYYYYMM1] [varchar](6) NULL

    ) ON [PRIMARY]

    GO

    INSERT INTO #mytable (TransactionYYYYMM1, Product)

    SELECT '201108', 'B' UNION ALL

    SELECT '201108', 'C' UNION ALL

    SELECT '201106', 'A' UNION ALL

    SELECT '201101', 'C' UNION ALL

    SELECT '201103', 'A' UNION ALL

    SELECT '201101', 'A' UNION ALL

    SELECT '201101', 'A' UNION ALL

    SELECT '201104', 'B' UNION ALL

    SELECT '201107', 'A' UNION ALL

    SELECT '201107', 'B' UNION ALL

    SELECT '201103', 'C' UNION ALL

    SELECT '201112', 'C' UNION ALL

    SELECT '201112', 'A' UNION ALL

    SELECT '201111', 'A' UNION ALL

    SELECT '201103', 'A' UNION ALL

    SELECT '201110', 'A' UNION ALL

    SELECT '201103', 'A' UNION ALL

    SELECT '201106', 'A' UNION ALL

    SELECT '201108', 'A' UNION ALL

    SELECT '201110', 'A'

    SELECT *

    FROM ( SELECT Product ,

    [201101] ,

    [201102] ,

    [201103] ,

    [201104] ,

    [201105] ,

    [201106] ,

    [201107] ,

    [201108] ,

    [201109] ,

    [201110] ,

    [201111] ,

    [201112]

    FROM ( SELECT Product,

    [TransactionYYYYMM1]

    FROM #mytable

    GROUP BY Product, TransactionYYYYMM1 WITH ROLLUP

    ) AS source PIVOT ( COUNT(TransactionYYYYMM1) FOR [TransactionYYYYMM1] IN ( [201101],

    [201102],

    [201103],

    [201104],

    [201105],

    [201106],

    [201107],

    [201108],

    [201109],

    [201110],

    [201111],

    [201112] ) ) AS pvt

    ) T1

    ORDER BY 1

  • Will this do the cut?

    ; WITH CTE AS

    (

    SELECT Product

    ,[TransactionYYYYMM1]

    ,Ct = COUNT(*)

    FROM #mytable

    GROUP BY Product, TransactionYYYYMM1

    UNION ALL

    SELECT 'TOTAL' Product

    , [TransactionYYYYMM1]

    , Ct = COUNT(*)

    FROM #mytable

    GROUP BY TransactionYYYYMM1

    --ORDER BY Product , [TransactionYYYYMM1]

    )

    SELECT C.Product

    ,[201101] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201101' THEN C.Ct ELSE 0 END)

    ,[201102] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201102' THEN C.Ct ELSE 0 END)

    ,[201103] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201103' THEN C.Ct ELSE 0 END)

    ,[201104] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201104' THEN C.Ct ELSE 0 END)

    ,[201105] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201105' THEN C.Ct ELSE 0 END)

    ,[201106] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201106' THEN C.Ct ELSE 0 END)

    ,[201107] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201107' THEN C.Ct ELSE 0 END)

    ,[201108] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201108' THEN C.Ct ELSE 0 END)

    ,[201109] = MAX( CASE WHEN C.TransactionYYYYMM1 = '201109' THEN C.Ct ELSE 0 END)

    ,[2011010] = MAX( CASE WHEN C.TransactionYYYYMM1 = '2011010' THEN C.Ct ELSE 0 END)

    ,[2011011] = MAX( CASE WHEN C.TransactionYYYYMM1 = '2011011' THEN C.Ct ELSE 0 END)

    ,[2011012] = MAX( CASE WHEN C.TransactionYYYYMM1 = '2011012' THEN C.Ct ELSE 0 END)

    FROM CTE C

    GROUP BY C.Product

    ORDER BY C.Product

  • Just to make it a bit more flexible so that you don't have to change the code every year, try the following against the test data you provided. I also added in a "Total" column and that automatically produces a "Grand Total", as well (heh... you just know they're going to ask for such things).

    --===== Declare some obviously named variables

    DECLARE @Year DATETIME, --@Year could be a parameter for a stored procedure

    @sql VARCHAR(MAX);

    --===== Preset the variables

    SELECT @Year = '2011';

    SELECT @sql = ' SELECT Product = CASE WHEN GROUPING(Product) = 0 THEN Product ELSE ''Total'' END,'

    + CHAR(10);

    --===== Build the dynamic SELECT list

    WITH

    cte12(N) AS

    (

    SELECT 0 UNION ALL SELECT 1 UNION ALL SELECT 2 UNION ALL SELECT 3 UNION ALL

    SELECT 4 UNION ALL SELECT 5 UNION ALL SELECT 6 UNION ALL SELECT 7 UNION ALL

    SELECT 8 UNION ALL SELECT 9 UNION ALL SELECT 10 UNION ALL SELECT 11

    ),

    cteDateParts AS

    (

    SELECT N,

    ColName = QUOTENAME(CONVERT(CHAR(6),DATEADD(mm,N,@Year),112)),

    TDate = QUOTENAME(CONVERT(CHAR(6),DATEADD(mm,N,@Year),112),'''')

    FROM cte12

    )

    SELECT @sql = @sql + SPACE(8) + ColName

    + '=SUM(CASE WHEN TransactionYYYYMM1=' + TDate + ' THEN 1 ELSE 0 END),'

    + CHAR(10)

    FROM CteDateParts

    ORDER BY N

    ;

    --===== Build the rest of the dynamic SQL

    SELECT @sql = @sql + SPACE(8) + 'Total = COUNT(*)

    FROM #MyTable

    GROUP BY Product WITH ROLLUP;'

    ;

    --===== Print the Dynamic SQL so we can see what it looks like.

    -- This section may be commented out for production.

    -- PRINT @sql

    ;

    --===== Produce the report

    EXEC (@SQL)

    ;

    --Jeff Moden


    RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
    First step towards the paradigm shift of writing Set Based code:
    ________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

    Change is inevitable... Change for the better is not.


    Helpful Links:
    How to post code problems
    How to Post Performance Problems
    Create a Tally Function (fnTally)

  • Thank you both....they work beautifully!

Viewing 8 posts - 1 through 7 (of 7 total)

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