Yes I've tried the code. My requirement is that user inputs Year and Month & prices are shown date-wise in columns for that year and month, with first column as CompetitorID. I want my result like:
Competitors | day1 | day2 | day3 | day4 ..............|day31 ================================================================ competitor 1| Price | Price | price |开发者_运维百科 price..............|price competitor 2| Price | Price | price | price..............|price competitor 3| Price | Price | price | price..............|price competitor 4| Price | Price | price | price..............|price
My Table structure is:
COMPETITORDETAIL (ID, CompetitorID, oDate, Price)
This is a lot easier. I wrote a stored proc named pivot_query that makes PIVOT a lot easier to use for SQL Server 2005+. The source for the proc is here, some examples how to use it are here.
For your code example:
create table Competitors
(
CompetitorId integer identity,
Name varchar(30)
)
insert into Competitors values ('Bobs Discount Emporium')
go
insert into Competitors values ('Joes Really Cheap Crap')
go
create table CompetitorDetail
(
Id integer identity,
CompetitorId integer,
oDate datetime,
Price decimal(12,3)
)
insert into CompetitorDetail values (1, getdate()-10, 10.00)
go
insert into CompetitorDetail values (1, getdate()-10, 10.00)
go
insert into CompetitorDetail values (1, getdate()-10, 10.00)
go
insert into CompetitorDetail values (1, getdate()-10, 10.00)
go
insert into CompetitorDetail values (1, getdate()-8, 11.00)
go
insert into CompetitorDetail values (1, getdate()-8, 11.00)
go
insert into CompetitorDetail values (1, getdate()-8, 11.00)
go
insert into CompetitorDetail values (1, getdate()-8, 11.00)
go
insert into CompetitorDetail values (1, getdate()-6, 12.00)
go
insert into CompetitorDetail values (1, getdate()-6, 12.00)
go
insert into CompetitorDetail values (1, getdate()-6, 12.00)
go
insert into CompetitorDetail values (1, getdate()-2, 13.00)
go
insert into CompetitorDetail values (1, getdate()-2, 13.00)
go
insert into CompetitorDetail values (1, getdate()-2, 13.00)
go
insert into CompetitorDetail values (2, getdate()-10, 14.00)
go
insert into CompetitorDetail values (2, getdate()-10, 14.00)
go
insert into CompetitorDetail values (2, getdate()-10, 14.00)
go
insert into CompetitorDetail values (2, getdate()-10, 14.00)
go
insert into CompetitorDetail values (2, getdate()-8, 15.00)
go
insert into CompetitorDetail values (2, getdate()-8, 15.00)
go
insert into CompetitorDetail values (2, getdate()-8, 15.00)
go
insert into CompetitorDetail values (2, getdate()-8, 15.00)
go
insert into CompetitorDetail values (2, getdate()-6, 16.00)
go
insert into CompetitorDetail values (2, getdate()-6, 16.00)
go
insert into CompetitorDetail values (2, getdate()-6, 16.00)
go
insert into CompetitorDetail values (2, getdate()-2, 18.00)
go
insert into CompetitorDetail values (2, getdate()-2, 18.00)
go
insert into CompetitorDetail values (2, getdate()-2, 18.00)
go
declare @mySQL varchar(MAX)
set @mySQL = '
select
c.Name,
right(cast(month(cd.oDate) + 100 as varchar(3)),2) + ''_'' + right(cast(day(cd.oDate) + 100 as varchar(3)),2) mon_day,
cd.Price
from
Competitors c
JOIN CompetitorDetail cd
on (cd.CompetitorId = c.CompetitorId )
';
exec pivot_query @mySQL, 'Name', 'Mon_Day', 'max(Price) MaxP,min(Price) MinP'
which results in:
Name 01_09_MaxP 01_09_MinP 01_11_MaxP 01_11_MinP 01_13_MaxP 01_13_MinP 01_17_MaxP 01_17_MinP
------------------------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------
Bobs Discount Emporium 10.000 10.000 11.000 11.000 12.000 12.000 13.000 13.000
Joes Really Cheap Crap 14.000 14.000 15.000 15.000 16.000 16.000 18.000 18.000
Hope that helps!
If you'r working with Microsoft SQL-Server there is a pivot function. But the columns are defined statically, so it's use is limited. (See http://technet.microsoft.com/en-us/library/ms177410.aspx)
There are solutions with dynamic SQL, but i solved this in the code eventually.
For SQL Server 2005, 2008
The table and some data to test
CREATE TABLE CompetitorDetail
(
ID int
,CompetitorID int
,oDate datetime
,Price decimal(19, 4)
)
INSERT INTO CompetitorDetail
( ID, CompetitorID, oDate, Price )
SELECT 1, 1, '2010-01-01', 100.0 UNION
SELECT 2, 1, '2010-01-02', 110.0 UNION
SELECT 3, 1, '2010-01-03', 99.0 UNION
SELECT 4, 2, '2010-01-01', 102.2 UNION
SELECT 5, 2, '2010-01-02', 112.2 UNION
SELECT 6, 2, '2010-01-03', 99.2 UNION
SELECT 7, 3, '2010-01-01', 100.3 UNION
SELECT 8, 3, '2010-01-02', 110.3 UNION
SELECT 9, 3, '2010-01-03', 99.3 ;
Start of the period and number of days
/* First day of the peroid */
DECLARE @StartDate datetime
,@NumberOfDays int
SET @StartDate = '2010-01-01'
SET @NumberOfDays = 31
Dynamic columns = dynamic sql
/* Table to compose dynamic query */
DECLARE @qw TABLE
(
id int IDENTITY(1, 1)
,txt nvarchar(500)
)
/* Start composing dynamic query */
INSERT INTO @qw ( txt ) VALUES ( 'SELECT' )
INSERT INTO @qw ( txt ) VALUES ( 'CompetitorID' )
Continue composing the dynamic query
/* Helpers */
DECLARE
@dte datetime
,@str varchar(10)
,@i int
/* Compose dynamic query */
SET @i = 0
WHILE @i < @NumberOfDays
BEGIN
SET @dte = DATEADD(d, @i, @StartDate)
SET @str = CONVERT(varchar(10), @dte, 121)
INSERT INTO @qw ( txt )
SELECT ',MAX(CASE oDate WHEN ''' + @str + ''' THEN Price ELSE NULL END) AS [' + @str + ']'
SET @i = @i + 1
END
/* Finish the dynamic query */
INSERT INTO @qw (txt) VALUES ( 'FROM CompetitorDetail' )
INSERT INTO @qw (txt) VALUES ( 'GROUP BY CompetitorID' )
INSERT INTO @qw (txt) VALUES ( 'ORDER BY CompetitorID' )
Concatenate into a variable and execute
/* Create a variable with dynamic sql*/
DECLARE @exe nvarchar(4000)
SET @exe=''
SELECT @exe = @exe + txt + ' ' FROM @qw ORDER BY id
/* execute dynamic sql */
EXEC sp_executesql @exe
Returns
CompetitorID 2010-01-01 2010-01-02 2010-01-03 2010-01-04 ... 2010-01-31
------------ ---------- ---------- ---------- ---------- ... ----------
1 100.0000 110.0000 99.0000 NULL ... NULL
2 102.2000 112.2000 99.2000 NULL ... NULL
3 100.3000 110.3000 99.3000 NULL ... NULL
CASE
WHEN SQL Server 2005 OR 2008 THEN Use Pivot
WHEN Oracle THEN Use LIKE MAX(Decode(Day=1,Data,0) as Day1 and GROUP BY Day
WHEN SQL Server 2000 THEN Use LIKE MAX(CASE WHEN Day = 1 THEN Data ELSE 0 End) As Day1 and GROUP BY Day
END
精彩评论