问题
I have the following dataset
Date Field1 Col1 Col2 Col3
2012/07/02 Customer1 CL DS RT
2012/07/03 Customer1 DS RT 700
2012/07/04 Customer1 DS RT 700
2012/07/02 Customer2 CL DS RT
2012/07/03 Customer2 DS RT 1500
2012/07/04 Customer2 DS RT 1500
2012/07/02 Customer3 CL DS RT
2012/07/03 Customer3 DS RT 6000
2012/07/04 Customer3 DS RT 6000
2012/07/02 Customer4 CL RC RT
2012/07/03 Customer4 RC RT 4900
2012/07/04 Customer4 RC RT 4900
I'd like output as follows:
Field1 2012/07/02 2012/07/03 2012/07/04
Col1 Customer1 CL DS DS
Col2 Customer1 DS RT RT
Col3 Customer1 RT 700 700
Col1 Customer2 CL DS DS
Col2 Customer2 DS RT RT
Col3 Customer2 RT 1500 1500
Col1 Customer3 CL DS DS
Col2 Customer3 DS RT RT
Col3 Customer3 RT 6000 6000
Col1 Customer4 CL RC RC
Col2 Customer4 RC RT RT
Col3 Customer4 RT 4900 4900
The problem is also that I have an unfixed amount of Customers (Field1) & unfixed amount of Dates.
回答1:
What you are trying to do is known as a PIVOT. There are two ways that these can be done, either with a static pivot where you hard-code all of the values to transform or a dynamic pivot where the values were determined at execution.
Static Version (See SQL Fiddle with Demo):
select *
from
(
select dt, field1, col, value
from yourTable
unpivot
(
value for col in (col1, col2, col3)
) u
)x1
pivot
(
min(value)
for dt in ([2012-07-02], [2012-07-03], [2012-07-04])
) p
Dynamic Pivot (See SQL Fiddle with Demo):
DECLARE @colsUnpivot AS NVARCHAR(MAX),
@query AS NVARCHAR(MAX),
@colsPivot as NVARCHAR(MAX)
select @colsUnpivot = stuff((select ','+quotename(C.name)
from sys.columns as C
where C.object_id = object_id('yourTable') and
C.name like 'col%'
for xml path('')), 1, 1, '')
select @colsPivot = STUFF((SELECT distinct ', ' + QUOTENAME(convert(char(10), dt, 120))
from yourTable
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
set @query =
'SELECT col, field1, ' + +@colsPivot+ '
FROM
(
select dt, field1, col, value
from yourTable
unpivot
(
value for col in ('+ @colsUnpivot+')
) u
)x1
pivot
(
min(value)
for dt in ('+@colsPivot+')
) p '
exec(@query)
The dynamic pivot is a great option when you do not know the number of items that you need to convert into columns. Both will yield the same results.
Edit #1, if you want the date columns in a specific order - desc, etc, then you will use the following to get the dates (See SQL Fiddle with Demo):
select @colsPivot = STUFF((SELECT ', ' + QUOTENAME(convert(char(10), dt, 120))
from yourTable
group by dt
ORDER by dt desc
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
回答2:
You don't need to do it without aggregates, but you do need to use Dynamic SQL to do it. Search for "SQL Dynamic Pivot" and you'll find plenty of posts about it. Like: SQL Server dynamic PIVOT query?
But better would be to do this in the client, as SQL's clients prefer to know what columns are being produced ahead of time.
回答3:
You must use SQL Server Pivot query for your question.
You Can Use Below Query :
CREATE TABLE mytable(date DATE, Field1 NVARCHAR(50), col1 NVARCHAR(30), col2 NVARCHAR(30),col3 NVARCHAR(30))
INSERT INTO mytable(date,Field1,col1,col2,col3)
VALUES('2012/07/02','Customer1','CL','DS','RT')
...
SELECT Col, Field1, [2012/07/02],[2012/07/03],[2012/07/04]
FROM (SELECT Field1, 'Col1' AS Col, Date, Col1 AS ColVal
FROM MyTable
UNION ALL
SELECT Field1, 'Col2' , Date, Col2 AS ColVal
FROM MyTable
UNION ALL
SELECT Field1, 'Col3' , Date, Col3 AS Col
FROM MyTable
)AS P
PIVOT (MIN(Colval) FOR Date IN ([2012/07/02],[2012/07/03],[2012/07/04])) AS Pvt
ORDER BY Field1, Col
来源：https://stackoverflow.com/questions/12140768/transpose-rows-and-columns-without-aggregate