Hi Erick,

Thanks for the reply and sorry, my fault, wasn't clear enough. I was
wondering if there was a way to remove terms that would always be zero
(because the term came from a document that didn't match the filter query).

Here's an example. I have a bunch of documents with fields 'manufacturer'
and 'location'. If I set my filter query to "manufacturer = Sony" and all
Sony documents had a value of 'Florida' for location, then I want 'Florida'
NOT to show up in my facet field results. Instead, it shows up with a count
of zero (and it'll always be zero because of my filter query).

Using mincount = 1 doesn't solve my problem because I don't want it to hide
zeroes that came from documents that actually pass my filter query.

Does that make more sense?


On Thu, Nov 21, 2013 at 4:36 PM, Erick Erickson <erickerick...@gmail.com>wrote:

> That's what faceting does. The facets are only tabulated
> for documents that satisfy they query, including all of
> the filter queries and anh other criteria.
>
> Otherwise, facet counts would be the same no matter
> what the query was.
>
> Or I'm completely misunderstanding your question...
>
> Best,
> Erick
>
>
> On Thu, Nov 21, 2013 at 4:22 PM, Luis Lebolo <luis.leb...@gmail.com>
> wrote:
>
> > Hi All,
> >
> > Is it possible to perform a facet field query on a subset of documents
> (the
> > subset being defined via a filter query for instance)?
> >
> > I understand that facet pivoting might work, but it would require that
> the
> > subset be defined by some field hierarchy, e.g. manufacturer -> price
> (then
> > only look at the results for the manufacturer I'm interested in).
> >
> > What if I wanted to define a more complex subset (where the name starts
> > with A but ends with Z and some other field is greater than 5 and yet
> > another field is not 'x', etc.)?
> >
> > Ideally I would then define a "facet field constraining query" to include
> > only terms from documents that pass this query.
> >
> > Thanks,
> > Luis
> >
>

Reply via email to