Question

请找到需要转置的表格（OututTable）。这里的QuestionID是通过连接两个值形成的 - [问题：AnswerID]

refID | SessionID | QuestionID     | AnswerValue
9000  | 205545715 | [4907]         | Good morning
12251 | 205543469 | [10576:16307]  | 3
12255 | 205543469 | [10907:17001]  | 4
13157 | 205543703 | [10576:16307]  | 3
14387 | 205543493 | [10907:17001]  | 2
14389 | 205543493 | [10911:17007]  | 3

预期输出每个SessionID应该有一行，而列数是动态的

SessionID | [4097]         | [10576:16307] | [10907:17001] | [10911:17007]
205545715 |Good morning    |               |               |
205543469 |                |    3          |   4           |
205543703 |                |    3          |               |
205543493 |                |               |   2           |   3

我有上述格式的输出，但只插入了NULL值而不是Answer值

我在想列名可能不匹配。任何帮助都会很棒！请告诉我。

代码：

    set @Questions = (STUFF((SELECT distinct ',[' + cast(i.SessionID as varchar(20))  + ']'          
            FROM OutputTable i
            FOR XML PATH(''), TYPE).value('.','VARCHAR(max)'), 1, 1, ''))

    print @Questions



    set @SQLQuery = 'select QuestionID,'+ @Questions +' from '+'('+ 'select SessionID,QuestionID,AnswerValue from OutputTable '+ ') p '+ 'PIVOT'+ '('+'max(Answervalue)'+'FOR p.SessionID IN ('+ @Questions +')' +') as pvt'

Answer 1

好问题！问题在于QuestionID中的括号。虽然这些对于Pivot Column Aliases是必需的，但它们不能用作字符串过滤器。

代码示例还会为期望的输出切换QuestionID和SessionID。

此代码将返回预期输出，排序略有不同。此处创建临时表以模拟OutputTable对象。这需要使用数据库表进行切换。

declare
    @Questions varchar(max),
    @SQLQuery varchar(max)

create table #OutputTable
    (
    refID int,
    SessionID int,
    QuestionID varchar(50),
    AnswerValue varchar(50)
    )
insert into #OutputTable
    values
        (9000,205545715,'[4907]','Good morning'),
        (12251,205543469,'[10576:16307]','3'),
        (12255,205543469,'[10907:17001]','4'),
        (13157,205543703,'[10576:16307]','3'),
        (14387,205543493,'[10907:17001]','2'),
        (14389,205543493,'[10911:17007]','3')

set @Questions = (STUFF((SELECT distinct ',' + cast(i.QuestionID as varchar(20))          
    FROM #OutputTable i
    FOR XML PATH(''), TYPE).value('.','VARCHAR(max)'), 1, 1, ''))

set @SQLQuery = '
    select SessionID,'+ @Questions +'
    from    (
            select
                SessionID,
                replace(replace(QuestionID,''['',''''),'']'','''') QuestionID,
                AnswerValue
            from #OutputTable
            ) p 
        PIVOT (
              max(Answervalue)
              FOR p.QuestionID IN ('+ @Questions +')
              ) as pvt
    order by SessionID desc'

exec(@SQLQuery)

SQL Server Pivot正在插入NULL值

1 个答案: