Question

所以我有一个看起来像这样的表：

 User_ID |   Job_Func 
 --------------------
 Adams   |   DEV
 Adams   |   NET_ENG
 Brice   |   QA
 Cohen   |   DEV
 Cohen   |   NET_ENG
 Cohen   |   SUPERVISOR

它应该看起来像这样

 UserName  |     Developer    |  NetworkEngineer  |   QA    |  Supervisor
 -------------------------------------------------------------------------
 Adams     |        1         |          1        |    0    |      0
 Brice     |        0         |          0        |    1    |      0
 Cohen     |        1         |          1        |    0    |      1

我已经阅读了Pivots和动态Pivot，但是这个概念使我有些失望。我不需要在Job_Func上求和，而是将其放入每个用户只有一行的存储桶中，并吐出自定义列名。

我知道只需要列出4个工作职能（将来可能是5个）。

出于报告目的，最终会将其显示出来。

对此有任何帮助，或者只是向我指出正确的方向都是有用的。

Answer 1

简便的方法是使用分组，求和和大小写-像这样：

SELECT  UserName,
    SUM(CASE WHEN Job_Func = 'DEV' THEN 1 ELSE 0 END) AS Developer,
    SUM(CASE WHEN Job_Func = 'NET_ENG' THEN 1 ELSE 0 END) AS NetworkEngineer,
    SUM(CASE WHEN Job_Func = 'QA' THEN 1 ELSE 0 END) AS  QA,
    SUM(CASE WHEN Job_Func = 'SUPERVISOR' THEN 1 ELSE 0 END) AS  Supervisor
FROM TABLE_NAME
GROUP BY UserName

如果即使列出了一项以上的职务，即使您只希望1，也可以使用MAX代替SUM

Answer 2

您可以使用ANSI方法来执行此操作，但是它有一些手动工作

SELECT DISTINCT
   USER_ID,
   (SELECT count(*) FROM TB1 WHERE TB1.JOB_FUNC = 'DEV' AND T.USER_ID = TB1.USER_ID) as Developer,
   (SELECT count(*) FROM TB1 WHERE TB1.JOB_FUNC = 'NET_ENG' AND T.USER_ID = TB1.USER_ID) as NetworkEngineer,
   (SELECT count(*) FROM TB1 WHERE TB1.JOB_FUNC = 'QA' AND T.USER_ID = TB1.USER_ID) as QA,
   (SELECT count(*) FROM TB1 WHERE TB1.JOB_FUNC = 'SUPERVISOR' AND T.USER_ID = TB1.USER_ID) as supervisor
 fROM TB1 T

或

SELECT  USER_ID,
    SUM(IFF(Job_Func = 'DEV', 1, 0)) AS Developer,
    SUM(IFF(Job_Func = 'NET_ENG', 1, 0)) AS NetworkEngineer,
    SUM(IFF(Job_Func = 'QA', 1, 0)) AS  QA,
    SUM(IFF(Job_Func = 'SUPERVISOR', 1, 0)) AS  Supervisor
FROM TB1
GROUP BY USER_ID

具有自定义列名称的T-SQL Pivot

2 个答案: