SQL - Remove duplicates after using a GROUP BY clause [duplicate] - sql

Question

This question already has answers here:
Get top 1 row of each group
(19 answers)
Closed 10 months ago.
Let's say I had two tables that looked like this:
Prod_SerialNo
Prod_TestOnAt
Prod_AccountNo
SN0001
2021-04-08
045678
SN0001
2021-01-14
067891
SN0001
2021-11-29
091234
SN0002
2022-01-19
045678
SN0002
2020-07-30
045678
SN0002
2022-03-30
012345
SN0003
2022-04-01
078912
SN0003
2022-02-19
089123
SN0003
2022-03-18
023456
S_AccountNo
S_AccountType
S_AccountName
012345
Homeowner
Adam Smith
023456
Homeowner
Jeremy Chan
034567
Manufacturer
Anne Hudson
045678
Distributor
Barney Jones
056789
Distributor
Jasmine Coleman
067891
Homeowner
Christian Lewis
078912
Distributor
Heather Ogden
089123
Homeowner
Stephen Gray
091234
Distributor
Antony Newman
The Prod Table tabulates specific product tests by what serial number was used, when the product was tested, and who tested it. (There are other things in the table, including a primary key not shown here)
The S Table is a list of subscribers with a variety of information about them. S_AccountNo is the parent to Prod_AccountNo.
I want to query when the last test was performed for each Serial Number and what account name it was that performed the test, but I don't want multiple results (duplicates) for the same serial number. I have tried the following code:
SELECT
Prod_SerialNo,
MAX(Prod_TestOnAt) AS "Last Time Tested",
S_AccountName
FROM Prod
INNER JOIN S ON S.S_AccountNo = Prod.Prod_AccountNo
GROUP BY Prod_SerialNo, S_AccountName
ORDER BY Prod_SerialNo
However, the query ends up outputting the same serial number on multiple rows even though I ask for the max TestOnAt date and I group by serial number. What am I getting wrong?

Answer 1

I think there is no need to use Group by you can get result with Row_Number like this:
SELECT
t.Prod_SerialNo,
t.Prod_TestOnAt AS "Last Time Tested",
t.S_AccountName
FROM (
SELECT
Prod_SerialNo,
Prod_TestOnAt,
S_AccountName,
ROW_NUMBER() OVER (PARTITION BY Prod_SerialNo ORDER BY Prod_TestOnAt DESC) rw
FROM Prod
INNER JOIN S ON S.S_AccountNo = Prod.Prod_AccountNo
) t
WHERE t.rw=1
ORDER BY t.Prod_SerialNo

Answer 2

You are grouping by Prod_SerialNo, S_AccountName so you will get duplicate Prod_SerialNo if multiple rows exist with that Prod_SerialNo and different S_AccountNames. You could do a MAX on Prod_TestOnAt and get that value with it's Prod_SerialNo, then join the result on the table to get your desired info using a subquery like so:
SELECT
p.[Prod_SerialNo],
max.[LastTimeTested],
s.[S_AccountName]
FROM PROD as p
INNER JOIN
(
SELECT
Prod_SerialNo,
MAX(Prod_TestOnAt) as [LastTimeTested]
FROM Prod
GROUP BY [Prod_SerialNo]
) as max
on max.[Prod_SerialNo] = p.[Prod_SerialNo] and max.[LastTimeTested] = p.[Prod_TestOnAt]
INNER JOIN S as s
ON s.[S_AccountNo] = p.[Prod_AccountNo]
ORDER BY p.[Prod_SerialNo]

Answer 3

If you don't like the solution using ROW_NUMBER an alternative is to use CROSS APPLY to identify the last Prod_TestOnAt and the associated Prod_AccountNo.
SELECT DISTINCT p.Prod_SerialNo, ca.Prod_TestOnAt, s.S_AccountName
FROM Prod p
CROSS APPLY (SELECT TOP 1 Prod_TestOnAt, Prod_AccountNo
FROM Prod
WHERE Prod_SerialNo = p.Prod_SerialNo
ORDER BY Prod_TestOnAt DESC) ca
INNER JOIN S ON S.S_AccountNo = ca.Prod_AccountNo