You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-user@lucene.apache.org by Ravi Gidwani <ra...@gmail.com> on 2011/05/09 04:25:13 UTC

upper limit to boost weight/value ?

Hello:

Is there any upper limit to the boost weight/value ? For example in the
following query :

&qf=exact_title^2000+exact_category^1900+exact_tags^1700

are these boost values acceptable and work as expected ?

Thanks,
~Ravi Gidwani

Re: upper limit to boost weight/value ?

Posted by Ravi Gidwani <ra...@gmail.com>.
Thanks Lance.

On Sun, May 8, 2011 at 7:34 PM, Lance Norskog <go...@gmail.com> wrote:

> There is no upper limit. These are floats. But they can be small too.
> Boost < 1 are 'under normal'.
>
> One radix sorting trick is to boost one field 1000 and another field
> 5. If the first field is a string facet, this gives each group of
> results in one long query. Lucene sorting does radix also, of course,
> but sometimes it is not the best tool.
>
>
>
>
> On 5/8/11, Ravi Gidwani <ra...@gmail.com> wrote:
> > Hello:
> >
> > Is there any upper limit to the boost weight/value ? For example in the
> > following query :
> >
> > &qf=exact_title^2000+exact_category^1900+exact_tags^1700
> >
> > are these boost values acceptable and work as expected ?
> >
> > Thanks,
> > ~Ravi Gidwani
> >
>
>
> --
> Lance Norskog
> goksron@gmail.com
>

Re: upper limit to boost weight/value ?

Posted by Lance Norskog <go...@gmail.com>.
There is no upper limit. These are floats. But they can be small too.
Boost < 1 are 'under normal'.

One radix sorting trick is to boost one field 1000 and another field
5. If the first field is a string facet, this gives each group of
results in one long query. Lucene sorting does radix also, of course,
but sometimes it is not the best tool.




On 5/8/11, Ravi Gidwani <ra...@gmail.com> wrote:
> Hello:
>
> Is there any upper limit to the boost weight/value ? For example in the
> following query :
>
> &qf=exact_title^2000+exact_category^1900+exact_tags^1700
>
> are these boost values acceptable and work as expected ?
>
> Thanks,
> ~Ravi Gidwani
>


-- 
Lance Norskog
goksron@gmail.com