In SQL server (2016), I want to convert 2 rows into 1 row with fields of both rows. I have this example:
在SQL server(2016)中,我想将2行转换为包含两行字段的1行。我有这个例子:
IF OBJECT_ID('tempdb.dbo.#MyTable') IS not NULL DROP TABLE #MyTable
CREATE TABLE #MyTable (
Direction varchar(1),
DateKey int,
ID varchar(8),
[Sessions] int
)
insert into #MyTable values('S', 20180301, 'ID123456', 46)
insert into #MyTable values('R', 20180301, 'ID123456', 99)
select * from #MyTable
Output:
输出:
Direction DateKey ID Sessions
S 20180301 ID123456 46
R 20180301 ID123456 99
The output I want is:
我想要的输出是:
DateKey ID S_Sessions R_Sessions
20180301 ID123456 46 99
So I tried this query but it won't work:
所以我尝试了这个查询,但它不起作用:
select DateKey,ID,
case Direction
when 'S' then [Sessions] as S_Sessions -- Incorrect syntax near the keyword 'as'.
else [Sessions] as R_Sessions
end
from #MyTable
Maybe I have to create an extra table, insert rows where direction='S' and then update the records with data where direction='R' but I wonder if there is a better way to do this.
也许我必须创建一个额外的表,插入direction ='S'的行,然后用direction ='R'的数据更新记录,但我想知道是否有更好的方法来做到这一点。
4 个解决方案
#1
0
CASE
in SQL is an expression that returns a single value. It cannot be used to control execution flow like in procedural languages.
SQL中的CASE是一个返回单个值的表达式。它不能像过程语言那样用于控制执行流程。
You can use conditional aggregation for this:
您可以使用条件聚合:
select DateKey, ID,
max(case Direction when 'S' then [Sessions] end) as S_Sessions,
max(case Direction when 'R' then [Sessions] end) as R_Sessions
from #MyTable
group by DateKey, ID
在这里演示
#2
3
use PIVOT
使用PIVOT
select *
from #MyTable
pivot
(
max(Sessions)
for Direction in ([S], [R])
) p
#3
1
assuming that your table contains the "pairs" S and R you can also use a self join
假设您的表包含“对”S和R,您还可以使用自联接
SELECT s.DateKey , s.ID , s.Sessions S_Sessions , r.Sessions R_Sessions
FROM #MyTable S
JOIN #MyTable R
ON s.ID = r.ID
AND s.DateKey = r.DateKey
WHERE S.Direction = 'S'
AND r.Direction = 'R'
#4
0
Try It ... It works for me . more variable more case and more left join table.
试试吧......它适合我。更多变量更多案例和更多左连接表。
select a.DateKey,a.ID,
(case a.Direction
when 'S' then a.Sessions
end) as S_Sessions,
(case b.Direction
when 'R' then b.Sessions
end) as R_Sessions
from mytable as a CROSS JOIN mytable as b ON a.ID=b.ID LIMIT 2,1
#1
0
CASE
in SQL is an expression that returns a single value. It cannot be used to control execution flow like in procedural languages.
SQL中的CASE是一个返回单个值的表达式。它不能像过程语言那样用于控制执行流程。
You can use conditional aggregation for this:
您可以使用条件聚合:
select DateKey, ID,
max(case Direction when 'S' then [Sessions] end) as S_Sessions,
max(case Direction when 'R' then [Sessions] end) as R_Sessions
from #MyTable
group by DateKey, ID
在这里演示
#2
3
use PIVOT
使用PIVOT
select *
from #MyTable
pivot
(
max(Sessions)
for Direction in ([S], [R])
) p
#3
1
assuming that your table contains the "pairs" S and R you can also use a self join
假设您的表包含“对”S和R,您还可以使用自联接
SELECT s.DateKey , s.ID , s.Sessions S_Sessions , r.Sessions R_Sessions
FROM #MyTable S
JOIN #MyTable R
ON s.ID = r.ID
AND s.DateKey = r.DateKey
WHERE S.Direction = 'S'
AND r.Direction = 'R'
#4
0
Try It ... It works for me . more variable more case and more left join table.
试试吧......它适合我。更多变量更多案例和更多左连接表。
select a.DateKey,a.ID,
(case a.Direction
when 'S' then a.Sessions
end) as S_Sessions,
(case b.Direction
when 'R' then b.Sessions
end) as R_Sessions
from mytable as a CROSS JOIN mytable as b ON a.ID=b.ID LIMIT 2,1