From: Andreas <maps.on@gmx.net>
To: Jasen Betts <jasen@xnet.co.nz>
Cc: pgsql-sql@postgresql.org
Subject: Re: Need help in grouping records
Date: Sun, 20 May 2012 11:08:25 +0200
Message-ID: <4FB8B489.3050701@gmx.net> (raw)
In-Reply-To: <jp9mvc$qmu$1@reversiblemaps.ath.cx>
References: <4FB7EEC6.1030004@gmx.net>
<jp9mvc$qmu$1@reversiblemaps.ath.cx>
Am 20.05.2012 05:04, schrieb Jasen Betts:
> On 2012-05-19, Andreas<maps.on@gmx.net> wrote:
>> Hi,
>>
>> I'm trying to fight against double entries in tables.
>> I got as far as I can find similar records with trigram string matching.
>> If I do this with a table compared to itself I get something like this:
>>
>> id_a, id_b
>> 3, 5
>> 3, 7
>> 5, 3
>> 5, 7
>> 7, 3
>> 7, 5
>> 11, 13
>> 13, 11
>>
>> so the records with the IDs 3, 5, 7 and 11, 13 are similar enough to
>> form a group.
>>
>> How would I get a list of record-IDs with a group-ID like this
>>
>> record_id, group_id
>> 3, 1
>> 5, 1
>> 7, 1
>> 11, 2
>> 13, 2
>>
>> Is there a way to get this by SQL ?
>
> select id_a,min(least(id_a,id_b)) from SOMETHING group by id_a
>
> close enough?
>
> or this: ?
>
> select id_a, rank() over order by g from
> ( select id_a,min(least(id_a,id_b)) as g from SOMETHING group by id_a ) as foo
>
>
>
Thanks :)
Thats by far more elegant as my approach with arrays I figured out in
the meantime.
I changed rank() to dense_rank() in your solution.
Functionally the 1st line does allready all the magic, though.
Great :)
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: pgsql-sql@postgresql.org
Cc: maps.on@gmx.net, jasen@xnet.co.nz
Subject: Re: Need help in grouping records
In-Reply-To: <4FB8B489.3050701@gmx.net>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox