Mailing-List: contact solr-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: solr-user@lucene.apache.org
MIME-Version: 1.0
In-Reply-To: <CAGva5M1uAph2OwxXRf6OwL=VOwmfBNEO0jbU37D4A0mgUgc9Sw@mail.gmail.com>
References: <CANzuQZQMnZgj6gh5BH+eHfi=KnC8BhiuhrQ6o1BLAiXxCO=4Cg@mail.gmail.com>
 <CAGva5M1uAph2OwxXRf6OwL=VOwmfBNEO0jbU37D4A0mgUgc9Sw@mail.gmail.com>
From: Amrit Sarkar <sarkaramrit2@gmail.com>
Date: Thu, 2 Nov 2017 03:15:05 +0530
Message-ID: <CAH8R_2YyJXoLJ+yBPKzyZiRYhbGu=iiiLzAJRUywzNziegXv2A@mail.gmail.com>
Subject: Re: Streaming Expression - cartesianProduct
To: solr-user@lucene.apache.org
Content-Type: multipart/alternative; boundary="94eb2c1cd14aedfea1055cf2c884"
archived-at: Wed, 01 Nov 2017 21:45:13 -0000

--94eb2c1cd14aedfea1055cf2c884
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Following Pratik's spot-on comment and not really related to your question,

Even the "partitionKeys" parameter needs to be specified the "over" field
while using "parallel" streaming.

Amrit Sarkar
Search Engineer
Lucidworks, Inc.
415-589-9269
www.lucidworks.com
Twitter http://twitter.com/lucidworks
LinkedIn: https://www.linkedin.com/in/sarkaramrit2

On Thu, Nov 2, 2017 at 2:38 AM, Pratik Patel <pratik@semandex.net> wrote:

> Roll up needs documents to be sorted by the "over" field.
> Check this for more details
> http://lucene.472066.n3.nabble.com/Streaming-Expressions-rollup-function-
> returning-results-with-duplicate-tuples-td4342398.html
>
> On Wed, Nov 1, 2017 at 3:41 PM, Kojo <rbsnkjmr@gmail.com> wrote:
>
> > Wrap cartesianProduct function with fetch function works as expected.
> >
> > But rollup function over cartesianProduct doesn=C2=B4t aggregate on a r=
eturned
> > field of the cartesianProduct.
> >
> >
> > The field "id_researcher" bellow is a Multivalued field:
> >
> >
> >
> > This one works:
> >
> >
> > fetch(reasercher,
> >
> >     cartesianProduct(
> >         having(
> >             cartesianProduct(
> >             search(schoolarship,zkHost=3D"localhost:9983",qt=3D"/export=
",
> > q=3D"*:*",
> > fl=3D"process, area, id_reasercher",sort=3D"process asc"),
> >             area
> >         ),
> >         eq(area, val(Anything))),
> >     id_reasercher),
> >     fl=3D"name, django_id",
> >     on=3D"id_reasercher=3Ddjango_id"
> > )
> >
> >
> > This one doesn=C2=B4t works:
> >
> > rollup(
> >
> >     cartesianProduct(
> >         having(
> >             cartesianProduct(
> >             search(schoolarship,zkHost=3D"localhost:9983",qt=3D"/export=
",
> > q=3D"*:*",
> > fl=3D"process, area, id_researcher, status",sort=3D"process asc"),
> >             area
> >         ),
> >         eq(area, val(Anything))),
> >     id_researcher),
> >     over=3Did_researcher,count(*)
> > )
> >
> > If I aggregate over a non MultiValued field, it works.
> >
> >
> > Is that correct, rollup doesn=C2=B4t work on a cartesianProduct?
> >
>

--94eb2c1cd14aedfea1055cf2c884--