pivot / unpivot

  • hi...pls help...

    i have a table with data as :

    emp_codepaydatebasichramedsplall ....

    a00131-01-141000020001001000

    a00231-01-142000020001001000

    a00331-01-141500020001001000

    a00431-01-141000020001001000

    a00531-01-141000020001001000

    a00631-01-141000020001001000

    a00131-12-131000020001001000

    a00231-12-132000020001001000

    a00331-12-131500020001001000

    a00431-12-131000020001001000

    a00531-12-131000020001001000

    a00631-12-131000020001001000

    .

    .

    ...............................................................

    AND NOW I NEED OUTPUT IN GIVEN FORMAT :

    31-01-1431-12-13

    BASIC7500075000

    HRA 1200012000

    MED 600 600

    SPLALL6000 6000

    NOTE : SUM OF BASIC OF ALL EMP IN A MONTH 31-01-14: 75000

  • Hi and welcome to ssc.

    There's an article for newcomers which you may find useful; http://www.sqlservercentral.com/articles/Best+Practices/61537/[/url]

    If you take a little time to prepare a sample data script, forum users will be encouraged to participate in your thread.

    This time I’ve prepared your sample data for you;

    -- Sample data setup

    SET DATEFORMAT DMY

    DROP TABLE #Sample

    CREATE TABLE #Sample (emp_code VARCHAR(4), paydate DATE, [basic] INT, hra INT, med INT, splall INT)

    INSERT INTO #Sample VALUES

    ('a001', '31-01-14', 10000, 2000, 100, 1000),

    ('a002', '31-01-14', 20000, 2000, 100, 1000),

    ('a003', '31-01-14', 15000, 2000, 100, 1000),

    ('a004', '31-01-14', 10000, 2000, 100, 1000),

    ('a005', '31-01-14', 10000, 2000, 100, 1000),

    ('a006', '31-01-14', 10000, 2000, 100, 1000),

    ('a001', '31-12-13', 10000, 2000, 100, 1000),

    ('a002', '31-12-13', 20000, 2000, 100, 1000),

    ('a003', '31-12-13', 15000, 2000, 100, 1000),

    ('a004', '31-12-13', 10000, 2000, 100, 1000),

    ('a005', '31-12-13', 10000, 2000, 100, 1000),

    ('a006', '31-12-13', 10000, 2000, 100, 1000)

    SELECT * FROM #Sample; -- check everything is ok (Stare & Compare)

    Here’s a solution which employs both CROSS APPLY VALUES [/url](a handy unpivot alternative championed by Dwain Camps) and CROSSTAB[/url] (championed by MVP Jeff Moden);

    -- Solution

    SELECT

    Category,

    [31-01-14] = SUM(CASE WHEN paydate = '31-01-14' THEN Amount END),

    [31-12-13] = SUM(CASE WHEN paydate = '31-12-13' THEN Amount END)

    FROM #Sample

    CROSS APPLY (

    VALUES

    ('basic', [basic]),

    ('hra', hra),

    ('med', med),

    ('splall', splall)

    ) x (Category, Amount)

    GROUP BY Category

    Test it rigorously. If there are any issues, post back, extending your sample data set if necessary to help explain your findings.

    “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

Viewing 2 posts - 1 through 1 (of 1 total)

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