Mailing-List: contact user-help@couchdb.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@couchdb.apache.org
Received-SPF: pass (athena.apache.org: domain of jchris@gmail.com designates
 209.85.212.177 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:sender:in-reply-to:references:date
         :x-google-sender-auth:message-id:subject:from:to:cc:content-type
         :content-transfer-encoding;
        b=GxvpKk9ckToIGP5+vlrV+VK9zDjSWwd6JzM8+Lg2Zb7EPCbK+mCxg14O3iA37RDnYU
         C/4y7lE0FoPkIDocUPOGqmMPzg59pzitMKYm+KR8QCn/5Y4K8zU9M6e6ShqJmzrtLJjk
         IrYx2nWpyz45aURp1xsAucO0Y1yMA0kQD8QGw=
MIME-Version: 1.0
Sender: jchris@gmail.com
In-Reply-To: <20090810195848.GD27797@uk.tiscali.com>
References: <41139fcb0908101135r64ead326k539a1b4eb60aed42@mail.gmail.com>
	 <20090810195848.GD27797@uk.tiscali.com>
Date: Mon, 10 Aug 2009 13:56:49 -0700
Message-ID: <e282921e0908101356u3178e427r451c712ba0878e5c@mail.gmail.com>
Subject: Re: sort by reduce value
From: Chris Anderson <jchris@apache.org>
To: user@couchdb.apache.org
Cc: Anand Chitipothu <anandology@gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

On Mon, Aug 10, 2009 at 12:58 PM, Brian Candler<B.Candler@pobox.com> wrote:
> On Tue, Aug 11, 2009 at 12:05:28AM +0530, Anand Chitipothu wrote:
>> I'm trying to think of couchdb equivalent of the following SQL query.
>>
>> SELECT author, count(*) =A0as count FROM changes GROUP BY author ORDER B=
Y count
>
> Being cheeky, I will point out that you can probably implement this in a
> _list view. But this is probably less efficient than just downloading the
> result set to the client and sorting it there.
>
> I'm assuming you know that you can get each author's count using a reduce
> view, and then query that with group=3Dtrue, so you'll already have calcu=
lated
> authors and counts. It's just the final sort by count which remains to be
> done client-side.
>

The reason why CouchDB can't do the sorting by value inside itself is
that on a multi node cluster, an individual couch can't know the
global ordering. We have a commitment to ensuring that CouchDB's API
remains invariant regardless of the size of the underlying cluster.

So you basically have to copy the cluster-wide reduce values to an
intermediate storage mechanism (which itself can be a cluster) before
you can sort them by value.

Cheers,
Chris

> Regards,
>
> Brian.
>


--=20
Chris Anderson
http://jchrisa.net
http://couch.io