Convert rows to columns considering multiple columns - sql

Question

I have current base table:
and I need to convert it to this:
I already have a dynamic SQL in place which is giving me below result for the LTD values:
But I am not able to create dynamically the columns for the Aggregationname based on the columns LTD and BOY and set the post fix to the column name.
Find my SQL below:
create table #tempIE
(
AGGREGATIONNAME varchar(2),
LTD decimal,
BOY decimal,
MONTH int
)
insert into #tempIE values ('XX', 50, 45, 00)
insert into #tempIE values ('XX', 150, 145, 01)
insert into #tempIE values ('XX', 300, 295, 02)
insert into #tempIE values ('YY', 25, 20, 00)
insert into #tempIE values ('YY', 50, 45, 01)
insert into #tempIE values ('YY', 75, 70, 02)
insert into #tempIE values ('ZZ', 500, 495, 00)
insert into #tempIE values ('ZZ', 600, 595, 01)
insert into #tempIE values ('ZZ', 700, 695, 02)
SELECT *
FROM #tempIE
DECLARE #cols AS NVARCHAR(MAX),
#query AS NVARCHAR(MAX);
SET #cols = STUFF((SELECT distinct ',' + QUOTENAME(c.AGGREGATIONNAME)
FROM #tempIE c
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
set #query = 'SELECT MONTH, ' + #cols + ' INTO ##tempIEnew from
(
select AGGREGATIONNAME
, LTD
--, BOY
, MONTH
from #tempIE
) x
pivot
(
max(LTD)
for AGGREGATIONNAME in (' + #cols + ')
) p '
;
execute(#query)
SELECT *
FROM ##tempIEnew
drop table #tempIE
drop table ##tempIEnew

Answer 1

As I mentioned in the comments, I much more prefer using conditional aggregation rather than the PIVOT operator, it's so much easier to work with as it's not got restrictive syntax.
With conditional aggregation, you can easily achieve a static pivot with the following:
SELECT IE.[MONTH],
MAX(CASE IE.AGGREGATIONNAME WHEN 'XX' THEN IE.LTD END) AS XX_LTD,
MAX(CASE IE.AGGREGATIONNAME WHEN 'YY' THEN IE.LTD END) AS YY_LTD,
/* ... */
MAX(CASE IE.AGGREGATIONNAME WHEN 'ZZ' THEN IE.BOY END) AS ZZ_BPY
FROM #tempIE IE
GROUP BY IE.[MONTH];
For a dynamic pivot, then you could achieve this with the following.
DECLARE #SQL nvarchar(MAX),
#CRLF nchar(2) = NCHAR(13) + NCHAR(10);
DECLARE #Delim nvarchar(3) = N',' + #CRLF
SELECT #SQL = N'SELECT IE.[MONTH],' + #CRLF +
STRING_AGG(N' MAX(CASE IE.AGGREGATIONNAME WHEN ' + QUOTENAME(IE.AGGREGATIONNAME,'''') + N' THEN IE.LTD END) AS ' + QUOTENAME(CONCAT(IE.AGGREGATIONNAME,'_LTD')),#Delim) WITHIN GROUP (ORDER BY IE.AGGREGATIONNAME) + N',' + #CRLF +
STRING_AGG(N' MAX(CASE IE.AGGREGATIONNAME WHEN ' + QUOTENAME(IE.AGGREGATIONNAME,'''') + N' THEN IE.BOY END) AS ' + QUOTENAME(CONCAT(IE.AGGREGATIONNAME,'_BOY')),#Delim) WITHIN GROUP (ORDER BY IE.AGGREGATIONNAME) + #CRLF +
N'FROM #tempIE IE' + #CRLF +
N'GROUP BY IE.[MONTH];'
FROM (SELECT DISTINCT AGGREGATIONNAME
FROM #tempIE) IE;
--PRINT #SQL;--Your best friend
EXEC sys.sp_executesql #SQL;
Note I assume you are using a fully supported version of SQL Server (as it's not noted you aren't) and so there's no need to FOR XML PATH and 2 passes of the table. I also switch to creating the query inline, as attempting to that with a variable #Columns is a real pain with conditional aggregation; I actually don't recommend that method either cause it's respective (like PIVOT).
db<>fiddle