t-sql secondary pivot on pivoted table - sql

Question

How would I pivot this table again to account for the difference in names?
What I have so far:
CREATE TABLE Temp
(
badge nvarchar(4)
,name nvarchar(31)
,Job nvarchar(4)
,KDA float
,Match int
)
INSERT INTO Temp
VALUES ('T996', 'Darrien', 'AP', 1.0, 20),
('T996', 'Mark', 'ADC', 2.8, 16),
('T996', 'Kevin', 'TOP', 5.0, 120)
SELECT badge, [AP_KDA], [AP_Match], [ADC_KDA], [ADC_Match], [TOP_KDA], [TOP_Match], [Person]
FROM (
SELECT badge, Col, Val
FROM (
SELECT badge, Job + '_KDA' AS Col, CAST(KDA AS nvarchar(31)) AS Val
FROM Temp
UNION ALL
SELECT badge, Job + '_Match' AS Col, CAST(Match AS nvarchar(31)) AS Val
FROM Temp
UNION ALL
SELECT badge, 'Person' AS Col, name AS Val
FROM Temp
) AS t
) AS tt
PIVOT (MIN(Val) FOR Col IN ([AP_KDA], [AP_Match], [ADC_KDA], [ADC_Match], [TOP_KDA], [TOP_Match], [Person])
) AS pvt
Which outputs:
badge AP_KDA AP_Match ADC_KDA ADC_Match TOP_KDA TOP_Match Person
1 T996 1 20 2.8 16 5 120 Darrien
I would like it to be formatted as:
badge AP_KDA AP_Match ADC_KDA ADC_Match TOP_KDA TOP_Match Person1 Person2 Person3
1 T996 1 20 2.8 16 5 120 Darrien Mark Kevin
I believe I'm close but the final Pivot is throwing me off.
Any help would be appreciated.
Thanks!

Answer 1

You can try to use condition aggregate function, MAX with CASE WHEN to make pivot.
CREATE TABLE Temp
(
badge nvarchar(4),
name nvarchar(31),
Job nvarchar(4),
KDA float,
Match int
)
INSERT INTO Temp VALUES
( 'T996' , 'Darrien' , 'AP' , 1.0, 20),
('T996' , 'Mark' , 'ADC' , 2.8 , 16),
( 'T996' , 'Kevin' , 'TOP' , 5.0 , 120)
Query 1:
SELECT badge,
MAX(CASE WHEN Job = 'AP' THEN KDA END) AP_KDA,
MAX(CASE WHEN Job = 'AP' THEN Match END) AP_Match,
MAX(CASE WHEN Job = 'ADC' THEN KDA END) ADC_KDA,
MAX(CASE WHEN Job = 'ADC' THEN Match END) ADC_Match,
MAX(CASE WHEN Job = 'TOP' THEN KDA END) TOP_KDA,
MAX(CASE WHEN Job = 'TOP' THEN Match END) TOP_Match,
MAX(CASE WHEN Job = 'AP' THEN name END) Person1 ,
MAX(CASE WHEN Job = 'ADC' THEN name END) Person2 ,
MAX(CASE WHEN Job = 'TOP' THEN name END) Person3
FROM TEMP
GROUP BY badge
Results:
| badge | AP_KDA | AP_Match | ADC_KDA | ADC_Match | TOP_KDA | TOP_Match | Person1 | Person2 | Person3 |
|-------|--------|----------|---------|-----------|---------|-----------|---------|---------|---------|
| T996 | 1 | 20 | 2.8 | 16 | 5 | 120 | Darrien | Mark | Kevin |

Answer 2

Please try this...
SELECT badge, [AP_KDA], [AP_Match], [ADC_KDA], [ADC_Match], [TOP_KDA], [TOP_Match],
[Person1],[Person2],[Person3]
FROM (
SELECT badge, Col, Val
FROM (
SELECT badge, Job + '_KDA' AS Col, CAST(KDA AS nvarchar(31)) AS Val
FROM Temp
UNION ALL
SELECT badge, Job + '_Match' AS Col, CAST(Match AS nvarchar(31)) AS
Val
FROM Temp
UNION ALL
SELECT badge, 'Person'+ CAST ((ROW_NUMBER() OVER (PARTITION BY badge
ORDER BY name DESC)) AS VARCHAR) AS Col, CAST(name AS VARCHAR) AS Val
FROM Temp
) AS t
) AS tt
PIVOT (MIN(Val) FOR Col IN ([AP_KDA], [AP_Match], [ADC_KDA], [ADC_Match],
[TOP_KDA], [TOP_Match], [Person1],[Person2],[Person3])
) AS pvt

Answer 3

If you want to dynamically create columns, meaning lets say you had a person4, you just cannot do it using PIVOT alone. You have to use For xml to transpose your data or another option is to use cursors, but I personally prefer For Xml. See this post for for xml