Re: slightly OT - cleaning up "dirty" keys?

From: bugbear <bugbear_at_trim_papermule.co.uk_trim>
Date: Fri, 03 Mar 2006 11:31:37 +0000
Message-ID: <4408291a$0$70321$ed2619ec@ptn-nntp-reader03.plus.net>

Joel Garry wrote:
> Is there any other information that might give a clue, ie, address?
> group by address having count(*) > 1
>
> Then all you have to worry about is wrong addresses :-)

Heh. That's the problem.

Different operators are FAILING to locate the original record, and are entering a new one. The address fields are just as "dirty" as th so-called (name) primary key.

All the data is a freakin' dirty mess.

I was hoping to cluster "likely"
primary-keys, so a human
could merge them, and also merge/fix the associated non-primary data.

BugBear Received on Fri Mar 03 2006 - 05:31:37 CST