在数据透视列中排序而不在结果中

Posted 2023-05-08

【中文标题】在数据透视列中排序而不在结果中

【英文标题】：Order by in Pivot Column not in Result

【发布时间】：2021-03-14 12:05:07

【问题描述】：

表-A

Productid	Brandname
1	C Brand
2	K Brand
3	A Brand

表-B

Productid	Rate	Slab
1	10	1
2	20	1
3	30	1
1	12	2
2	22	2
3	32	2

动态枢轴查询：-

DECLARE @cols AS NVARCHAR(MAX),
    @query  AS NVARCHAR(MAX)
select @cols = STUFF((SELECT distinct ',' + QUOTENAME(Brandname) from [Table-A] FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)'),1,1,'')
set @query = 'SELECT Slab,' + @cols + '
             from 
             (
                select a.BrandName, b.Rate,b.Slab from [Table-A] a inner join [Table-B] b on a.productid=b.productid 
            ) x
            pivot 
            (
                max(Rate)
                for Brandname in (' + @cols + ')
            ) p '

execute(@query)

结果：-

Slab	A Brand	C Brand	K Brand
1	30	10	20
2	32	12	22

但我需要使用 productid 订购 Brandname 列

预期结果：-

Slab	C Brand	K Brand	A Brand
1	10	20	30
2	12	22	32

【参考方案1】：

在为动态数据透视准备列列表时，使用order by productid并删除 distinct，您的问题将得到解决。

架构和插入语句：

 create table [Table-A](Productid   int,Brandname varchar(10));
 insert into [Table-A] values(1 ,'C Brand');
 insert into [Table-A] values(2 ,'K Brand');
 insert into [Table-A] values(3 ,'A Brand');
 
 create table [Table-B] (Productid  int, Rate   int , Slab int)
 insert into [Table-B] values(1 ,10 ,1);
 insert into [Table-B] values(2 ,20 ,1);
 insert into [Table-B] values(3 ,30 ,1);
 insert into [Table-B] values(1 ,12 ,2);
 insert into [Table-B] values(2 ,22 ,2);
 insert into [Table-B] values(3 ,32,2);

查询：

     DECLARE @cols AS NVARCHAR(MAX),
     @query  AS NVARCHAR(MAX)
 
    select @cols = STUFF((SELECT ',' + QUOTENAME(Brandname) from [Table-A] order by productid FOR XML PATH(''), TYPE
             ).value('.', 'NVARCHAR(MAX)'),1,1,'')
 
    
 set @query = 'SELECT Slab,' + @cols + '
              from 
              (
                 select a.BrandName, b.Rate,b.Slab from [Table-A] a inner join [Table-B] b on a.productid=b.productid 
             ) x
             pivot 
             (
                 max(Rate)
                 for Brandname in (' + @cols + ')
             ) p '
 
 execute(@query)

输出：

Slab	C Brand	K Brand	A Brand
1	10	20	30
2	12	22	32

如果您使用的是 SQL Server 2017 或更高版本，则 string_agg() 将是更智能、更快的选择。

 DECLARE @cols AS NVARCHAR(MAX),
     @query  AS NVARCHAR(MAX)
 
 select distinct @cols=STRING_AGG(quotename(brandname),',')  from [Table-A] 
 
     set @query = 'SELECT Slab,' + @cols + '
                  from 
                  (
                     select a.BrandName, b.Rate,b.Slab from [Table-A] a inner join [Table-B] b on a.productid=b.productid 
                 ) x
                 pivot 
                 (
                     max(Rate)
                     for Brandname in (' + @cols + ')
                 ) p '
     
     execute(@query)

输出：

Slab	C Brand	K Brand	A Brand
1	10	20	30
2	12	22	32

db小提琴here

【讨论】：

@Sathesh Kumar 如果您在此查询中遇到更多困难，请告诉我。

如果对您有帮助，请点击绿色检查以接受此答案。

【参考方案2】：

您希望列按特定顺序排列。根据您的数据，您似乎不需要子查询中的SELECT DISTINCT。但如果你这样做，你可以使用聚合来代替：

select @cols = STUFF((SELECT ',' + QUOTENAME(Brandname) 
                      FROM [Table-A]
                      GROUP BY BrandName
                      ORDER BY MIN(ProductId)
                      FOR XML PATH(''), TYPE
                     ).value('.', 'NVARCHAR(MAX)'
                            ), 1, 1, '')

但是，没有聚合：

select @cols = STUFF((SELECT ',' + QUOTENAME(Brandname) 
                      FROM [Table-A]
                      ORDER BY ProductId
                      FOR XML PATH(''), TYPE
                     ).value('.', 'NVARCHAR(MAX)'
                            ), 1, 1, '');

一旦@cols 的顺序正确，查询的其余部分应该就位。