數據透視表錯誤

Question

爲什麼，爲什麼？我是否收到錯誤：

「Msg 325，Level 15，State 1，Line 17 'PIVOT'附近的語法不正確您可能需要將當前數據庫的兼容性級別設置爲更高的值以啓用此功能。請參閱存儲過程sp_dbcmptlevel的幫助。「 對於此查詢？

WITH Offnet7 AS (
    SELECT disposition.dispositiondesc, interaction.dispositionid, DATEPART(wk,interaction.ibegintime) as iWeek 

    FROM interaction INNER JOIN 
      disposition ON interaction.reasonid = disposition.dispositionid 

    WHERE interaction.dispositionid = 10 and (reasonid = 20365 or reasonid = 20366 or reasonid = 11168) and 
      interaction.ibegintime >= '2013-1-1' and 
      interaction.ibegintime < '2014-1-1' 
) 

SELECT iWeek, dispositiondesc, count(iWeek) as 'OffnetCounts' 
FROM Offnet7 

Group by dispositiondesc, iWeek 

PIVOT 
(
    OffnetCounts 
    for iWeek in ([1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]) 
) AS counts 

編輯：「構建模式」 試圖使SQL小提琴，當我去給它哽咽（SQL Noob） 我從SQL管理工作室的對象瀏覽器中提取類型並複製了一些示例數據。但是，這是我嘗試：

CREATE TABLE interaction 
    ([reasonid] int, [dispositionid] int, [ibegintime] datetime) 
; 

INSERT INTO interaction 
    ([reasonid], [dispositionid], [ibegintime]) 
VALUES 
    (20366, 10, '2012-01-31 23:59:48.000'), 
    (20366, 10, '2012-02-07 14:03:01.000'), 
    (20366, 10, '2012-02-07 14:06:48.000'), 
    (20366, 10, '2012-02-13 21:44:10.000'), 
    (20366, 10, '2012-02-27 21:36:33.000') 
; 


CREATE TABLE disposition 
    ([dispositionid] int, [predefined] int, [dispositiondesc] varchar(64), [displayvalue] varchar(254)) 
; 

INSERT INTO disposition 
    ([dispositionid], [predefined], [dispositiondesc], [displayvalue]) 
VALUES 
(10, 1, 'TRANSFERRED OFFNET', 'TRANSFERRED OFFNET'), 
    (11168, 0, 'TAKEDA PASSWORD', 'TAKEDA PASSWORD'), 
    (15433, 0, 'Voice Mail - TAKEDAEMEA', 'Voice Mail - TAKEDAEMEA'), 
    (20365, 0, 'TAKEDA iPAD, iPhone or BlackBerry', 'TAKEDA iPAD, iPhone or BlackBerry'), 
    (20366, 0, 'TAKEDA Concur', 'TAKEDA Concur') 
; 

結論： 感謝所有幫助Bluefeet！

對於任何對此感興趣的人，如果我的DBA正確設置了SQL兼容性級別，他的第一個答案會起作用。想他的第一個答案之後，我得到了一個：

"Msg 102, Level 15, State 1, Line 19 Incorrect syntax near '('." 

因爲DBA沒有與支持PIVOT語句中的兼容性級別配置的SQL Server。

Answer 1

您的語法關閉。 PIVOT正在執行GROUP BY和聚合。在我看來，要使用：

WITH Offnet7 AS 
(
    SELECT disposition.dispositiondesc, 
    interaction.dispositionid, 
    DATEPART(wk,interaction.ibegintime) as iWeek 
    FROM interaction 
    INNER JOIN disposition 
    ON interaction.reasonid = disposition.dispositionid 
    WHERE interaction.dispositionid = 10 
    and (reasonid = 20365 or reasonid = 20366 or reasonid = 11168) 
    and interaction.ibegintime >= '2013-1-1' 
    and interaction.ibegintime < '2014-1-1' 
) 
SELECT dispositiondesc, 
    [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15] 
FROM Offnet7 
PIVOT 
(
    count(dispositionid) 
    for iWeek in ([1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]) 
) AS counts; 

見Demo

這將創建數據表與dispositionid的的每個星期的計數，由dispositiondesc分組。

編輯，這也可以使用聚合函數與CASE表達做到：

SELECT disposition.dispositiondesc, 
    sum(case when DATEPART(wk,interaction.ibegintime) = 1 then 1 else 0 end) [1], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 2 then 1 else 0 end) [2], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 3 then 1 else 0 end) [3], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 4 then 1 else 0 end) [4], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 5 then 1 else 0 end) [5], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 6 then 1 else 0 end) [6], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 7 then 1 else 0 end) [7], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 8 then 1 else 0 end) [8], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 9 then 1 else 0 end) [9], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 10 then 1 else 0 end) [10], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 11 then 1 else 0 end) [11], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 12 then 1 else 0 end) [12], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 13 then 1 else 0 end) [13], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 14 then 1 else 0 end) [14], 
    sum(case when DATEPART(wk,interaction.ibegintime) = 15 then 1 else 0 end) [15] 
FROM interaction 
INNER JOIN disposition 
    ON interaction.reasonid = disposition.dispositionid 
WHERE interaction.dispositionid = 10 
    and (reasonid = 20365 or reasonid = 20366 or reasonid = 11168) 
    and interaction.ibegintime >= '2013-1-1' 
    and interaction.ibegintime < '2014-1-1' 
group by disposition.dispositiondesc; 

見SQL Fiddle with Demo