我们如何使用PIVOT进行以下转换？

Question

我有以下

使用以下查询生成

declare @t table(Bucket varchar(50), [RollBack] int, [Stabilized] int, [RollForward] int,[Normalized] int)

insert into @t 
    select 'Bucket >6',0,0,100,0 union all
    select 'Bucket 1',0,0,200,0 union all
    select 'Bucket 2',0,0,100,0 union all
    select 'Bucket 3',0,0,100,0 union all
    select 'Bucket 4',0,0,100,0 union all
    select 'Bucket 5',0,0,100,0 union all
    select 'Bucket 6',0,0,100,0

select * from @t

我正在寻找下面的转换

到目前为止我的尝试是

select a as Activities,

    sum(case when b.Bucket='Bucket 0' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket 0],

    sum(case when b.Bucket='Bucket 1' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket 1],

    sum(case when b.Bucket='Bucket 2' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket 2],

    sum(case when b.Bucket='Bucket 3' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket 3],

    sum(case when b.Bucket='Bucket 4' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket 4],

    sum(case when b.Bucket='Bucket 5' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket 5],

    sum(case when b.Bucket='Bucket 6' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket 6],

    sum(case when b.Bucket='Bucket >6' then 
                    case when a='RollBack' then [RollBack] 
                         when a='Stabilized' then Stabilized
                         when a='RollForward' then RollForward
                         when a='Normalized' then Normalized                         
                    end else 0 end) as [Bucket >6]

    from (values ('RollBack'),('Stabilized'),('RollForward'),('Normalized')) t(a)
            cross join @t b
            group by a  

这给了我

Answer 1

我将使用make进行休息操作：

APPLY

但是，我不知道您的实际数据模型，所以我只用WITH t AS ( <aggergate query here> ) SELECT t.Activities, tt.vcol, tt.val FROM t CROSS APPLY ( VALUES ('Bucket 0', [Bucket 0]), . . . ) tt (col, val); 到apply数据来做。