I have the following syntax which created the below table:
SELECT [OPEN_DATE]
,[CODE]
,[ID]
,ROW_NUMBER() OVER (PARTITION BY [ID],[CODE] ORDER BY [OPEN_DATE],[CODE]) AS [RowOrder]
FROM [Table]
OPEN_DATE CODE ID RowOrder
6/8/2021 AA 11052 1
6/8/2021 AA 11052 2
6/8/2021 AA 11052 3
6/8/2021 AB 11052 1
6/8/2021 AB 11052 2
6/8/2021 AB 11052 3
7/15/2021 AC 89321 1
7/15/2021 AC 89321 2
7/15/2021 AC 89321 3
9/24/2022 AD 89321 1
9/24/2022 AD 89321 2
9/24/2022 AD 89321 3
I need to add column to number these in sets/groups like so:
OPEN_DATE CODE ID RowOrder SetOrder
6/8/2021 AA 11052 1 1
6/8/2021 AA 11052 2 1
6/8/2021 AA 11052 3 1
6/8/2021 AB 11052 1 2
6/8/2021 AB 11052 2 2
6/8/2021 AB 11052 3 2
7/15/2021 AC 89321 1 1
7/15/2021 AC 89321 2 1
7/15/2021 AC 89321 3 1
9/24/2022 AD 89321 1 2
9/24/2022 AD 89321 2 2
9/24/2022 AD 89321 3 2
Notice SetOrder increases by 1 when Code changes but ID stays the same, and then it resets to 1 when ID changes.
I tried using ROW_NUMBER again as well as DENSE_RANK and RANK, but none of them seem to work. Is there another function I haven't thought of?
Thank you.
You can use DENSE_RANK
for this
SELECT OPEN_DATE
,CODE
,ID
,ROW_NUMBER() OVER (PARTITION BY ID, CODE ORDER BY OPEN_DATE) AS RowOrder
,DENSE_RANK() OVER (PARTITION BY ID ORDER BY CODE) AS SetOrder
FROM [Table]
Note that CODE
in the ORDER BY
of RowOrder
makes no sense, as it's already in the PARTITION BY
.