sql top 1 based on group records - sql

Question

Table ProductMst (combination is unique - CodeNum, MainCatid, SubCatid)
id CodeNum MainCatid SubCatid Desc Qty
1 001 1 1 prod1 5
2 001 1 2 prod2 10
3 001 2 3 prod3 1
4 002 2 3 prod4 2
5 003 2 3 prod5 3
6 004 1 1 prod6 4
Sub Table SubCatMst
id name subcode
1 scat1 00
2 scat2 00
3 scat3 02
4 scat4 03
Desired Result
id CodeNum MainCatid SubCatid Desc Qty subcode
2 001 1 2 prod2 15 00
3 001 2 3 prod3 1 02
4 002 2 3 prod4 2 02
5 003 2 3 prod5 3 02
6 004 1 1 prod6 4 00
Basically, i wanted to group subcode if same and sum (Qty) based on subcode. And, take the TOP 1 records data as all columns based on MAX(Qty).
Summary: Please check first 2 records which combines.
Query Attempt:
select * from (
select A.*, B.subcode,
ROW_NUMBER() OVER( PARTITION BY A.CodeNum, A.MainCatid, B.subcode ORDER BY A.Qty desc) as row
from ProductMst A
inner join SubCatMst B on A.SubCatid=B.id
) as A where row<=1

Answer 1

Your question says:
i wanted to group subcode if same
If what is the same? I assume this means if codenum is the same.
Regardless of the exact field(s), the logic is the same. You can use window functions to aggregate the data and determine the row to choose:
select ps.*
from (select p.*, sc.subcode,
sum(qty) over (partition by codenum, subcode) as qty,
row_number() over (partition by codenum, subcode order by qty desc) as seqnum
from ProductMst p join
SubCatMst sc
on p.subcatid = sc.id
) ps
where seqnum = 1;