Using OVER() to Remove Duplicate Rows

Question

Using OVER() to Remove Duplicate Rows

Viewing 6 posts - 16 through 20 (of 20 total)

You must be logged in to reply to this topic. Login to reply

Mike McQuillan SSCertifiable Points: 6000 More actions · Answer 1

Thanks for the comment pdanes.

Indeed, as we all know there are several ways to achieve the same thing (as eduarrr and others have pointed out). There will always be views on what is "better". Glad the key point about the OVER clause resonated with you.

Cheers,

Mike.

Neil Burton SSC-Insane Points: 23009 More actions · Answer 2

Irfan.Baig (11/7/2016)
Hello Dear Mike,
I think we can also remove duplicates using simplest, easiest and better way without using temp table/s logic:
WITH CTE_InsurancePolicy AS (
SELECT PolicyId, FirstName, Surname, PolicyStartDate, PolicyEndDate
,ROW_NUMBER() OVER (PARTITION BY PolicyId, FirstName, Surname, PolicyStartDate, PolicyEndDate ORDER BY PolicyId) AS RowNumber
FROM dbo.InsurancePolicy
)
DELETE CTE_InsurancePolicy
WHERE RowNumber > 1;
Thanks
Irfan Baig.

That's brilliant. I've never seen that before and I found a use for it immediately. I'd got a CTE to create the row numbers but I didn't know you could DELETE directly from it.

On two occasions I have been asked, "Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?" ... I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question.
—Charles Babbage, Passages from the Life of a Philosopher

How to post a question to get the most help http://www.sqlservercentral.com/articles/Best+Practices/61537

thisisfutile Hall of Fame Points: 3668 More actions · Answer 3

Yev.d (11/8/2016)
Irfan.Baig (11/7/2016)
I think we can also remove duplicates using simplest, easiest and better way without using temp table/s logic:
I think the simplest in this case would be to truncate the table and re-run the import process.
😉

Simplest, yes. However, my pride is WAY to much in the way and I would definitely not go backward by starting over.:hehe:

My wife could share stories of long(er) car rides (before GPS's) because there was no way I was going to turn back and go the way we just came. There's another way...maybe even a shorter way...somewhere. lol

ken.trock SSCertifiable Points: 5147 More actions · Answer 4

Whoa, I didn't realize either that deleting from the CTE could delete from the original table. And you can insert/update too. Glad I didn't learn that lesson at an inopportune time!

This method of duplicate deletion sort of imposes a PK where there is none.

Ken

BWFC (11/10/2016)
Irfan.Baig (11/7/2016)
Hello Dear Mike,
I think we can also remove duplicates using simplest, easiest and better way without using temp table/s logic:
WITH CTE_InsurancePolicy AS (
SELECT PolicyId, FirstName, Surname, PolicyStartDate, PolicyEndDate
,ROW_NUMBER() OVER (PARTITION BY PolicyId, FirstName, Surname, PolicyStartDate, PolicyEndDate ORDER BY PolicyId) AS RowNumber
FROM dbo.InsurancePolicy
)
DELETE CTE_InsurancePolicy
WHERE RowNumber > 1;
Thanks
Irfan Baig.
That's brilliant. I've never seen that before and I found a use for it immediately. I'd got a CTE to create the row numbers but I didn't know you could DELETE directly from it.

akljfhnlaflkj SSC Guru Points: 76202 More actions · Answer 5

Good reminder how OVER works, thanks. I don't use it enough but when I do it can be a lifesaver.

rwoo_98 Grasshopper Points: 11 More actions · Answer 6

I've used over to create row numbers and to get the a certain row (example second row) within a group. However for getting distinct, this seems like overkill. Why not just use

select PolicyId, FirstName, Surname PolicyStartDate, PolicyEndDate

from dbo.InsurancePolicy

group by PolicyId, FirstName, Surname PolicyStartDate, PolicyEndDate

order by PolicyId