How to pivot by two columns into predefined table? - sql-server

Question

I've seen this question, this and this, however these are not what I want.
Please, do not close my question as it is not duplicate. It is really important to me.
I've managed to pivot a table, but this is not desired result:
The desired result looks like this:
My sample data is(it is just an example as OperatorX and OC columns should be 50 times):
DECLARE #OperatorPrice TABLE (ID INT NOT NULL, OperatorId INT NULL, Price
NUMERIC(18,3) NULL, FName VARCHAR(50) NULL)
INSERT INTO #OperatorPrice (
ID, OperatorId, Price, FName
)
VALUES
(226, 996, 22954,'Operator1')
, (266, 1016, 79011.2, 'Operator3')
, (112, 1029, 14869, 'Operator4')
, (93, 1031, 10568.96, 'Operator5')
DECLARE #TR TABLE
(
ID INT NULL ,
Operator1 DECIMAL(18,3) NULL, OC1 DECIMAL(18,3) NULL, Operator2 DECIMAL(18,3) NULL,
OC2 DECIMAL(18,3) NULL, Operator3 DECIMAL(18,3) NULL, OC3 DECIMAL(18,3) NULL,
Operator4 DECIMAL(18,3) NULL, OC4 DECIMAL(18,3) NULL, Operator5 DECIMAL(18,3) NULL,
OC5 DECIMAL(18,3) NULL
)
Example code:
INSERT #TR
(ID ,
Operator1, OC1, Operator2, OC2, Operator3, OC3, Operator4, OC4,
Operator5, OC5)
SELECT ID ,
Operator1, OC1, Operator2, OC2, Operator3, OC3, Operator4, OC4,
Operator5, OC5
FROM
(SELECT Price, id, FName
FROM #OperatorPrice) AS SourceTable
PIVOT
(
sum(Price)
FOR FName IN (Operator1, OC1, Operator2, OC2, Operator3, OC3,
Operator4, OC4, Operator5, OC5)
) AS PivotTable
SELECT * FROM #TR
How can I insert data into OC columns?

Answer 1

Perhaps something like this.
In your alias SOURCETABLE, we just add a UNION ALL of possible combinations for an ID with NULL values. In this case the MIN(ID) and values 1 - 50
Just be sure to
1) Define #TR with columns Operator1,OC1,..,Operator50,OC50 <<< OC# can be an INT
2) in the FOR Item IN(Operator1,OC1,..,Operator50,OC50)
Example -- Edit Corrected to allow for >9 operators
INSERT #TR
SELECT *
FROM (
Select A.ID
,B.*
From #OperatorPrice A
Cross Apply ( values (FName,Price)
,('OC'+replace(FName,'Operator',''),OperatorID)
) B (Item,Value)
Union All
Select ID=(select min(ID) From #OperatorPrice)
,B.*
From ( Select Top 50 N=Row_Number() Over (Order By (Select NULL)) From master..spt_values n1 ) A
Cross Apply ( values (concat('Operator',N),NULL)
,(concat('OC',N),NULL)
) B (Item,Value)
) AS SourceTable
PIVOT ( sum(Value) FOR Item IN (Operator1, OC1, Operator2, OC2, Operator3, OC3, Operator4, OC4, Operator5, OC5) ) AS PivotTable
Select * from #TR
Returns -- Notice Operator2

Answer 2

Using your sample tables and data this is pretty easy with conditional aggregation. Not really clear though how you determine which operator number is which. Hopefully you have something better than parsing numbers out of the values but who knows.
select op.ID
, Operator1 = max(case when convert(int, replace(FName, 'Operator', '')) = 1 then Price end)
, OC1 = max(case when convert(int, replace(FName, 'Operator', '')) = 1 then OperatorID end)
, Operator2 = max(case when convert(int, replace(FName, 'Operator', '')) = 2 then Price end)
, OC2 = max(case when convert(int, replace(FName, 'Operator', '')) = 2 then OperatorID end)
, Operator3 = max(case when convert(int, replace(FName, 'Operator', '')) = 3 then Price end)
, OC3 = max(case when convert(int, replace(FName, 'Operator', '')) = 3 then OperatorID end)
, Operator4 = max(case when convert(int, replace(FName, 'Operator', '')) = 4 then Price end)
, OC4 = max(case when convert(int, replace(FName, 'Operator', '')) = 4 then OperatorID end)
, Operator5 = max(case when convert(int, replace(FName, 'Operator', '')) = 5 then Price end)
, OC5 = max(case when convert(int, replace(FName, 'Operator', '')) = 5 then OperatorID end)
from #OperatorPrice op
cross apply
(
values
(1)
,(2)
,(3)
,(4)
)x(N)
group by op.ID