So, would the best way to be just to create an average column on my
item table, and then do facets against that?

Manish

On Aug 21, 5:44 am, Pat Allan <[email protected]> wrote:
> You're not wrong (although, should be an option on Riddle on the 0.9.9  
> branch) - however, SetSelect is not for attribute definition, but for  
> what (possibly computed) attributes get returned on search calls.  
> Also, I'm not sure if AVG gets calculated across a single MVA set, so  
> it might not do what you're requesting anyway.
>
> --
> Pat
>
> On 21/08/2009, at 10:18 AM, zaadjis wrote:
>
>
>
> > Actually, i think sphinx and the official Ruby API supports aggregates
> > (AVG, MIN, MAX, COUNT) since 0.9.9-rc2. But I don't think riddle/
> > thinking-sphinx does :(  Am i wrong?
>
> >http://sphinxsearch.com/docs/current.html#api-func-setselect
>
> > On Aug 20, 7:51 pm, Pat Allan <[email protected]> wrote:
> >> I've not tried this myself, but you'll need to do something along the
> >> lines of:
> >>    has "CAST(AVG(item_ratings.rating) as unsigned)", :as
> >> => :item_rating,
> >>      :facet => true, :type => :integer
> >>    has item_ratings(:id), :as => :item_ratings_ids # to ensure join
> >> exists
>
> >> --
> >> Pat
>
> >> On 20/08/2009, at 4:24 PM, mrmanishs wrote:
>
> >>> I want to getaverageratings for an item for facets, how do I do
> >>> this?
>
> >>> Example pseudo code:
> >>> Currently:
> >>>    has item_ratings(:rating), :as => :item_rating, :facet => true
>
> >>> Would llike something like:
> >>>    has item_ratings(:rating).average.to_i, :as
> >>> => :item_rating, :facet => true
>
> >>> Thanks!
>
> >>> Manish
>
>
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Thinking Sphinx" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/thinking-sphinx?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to