Hi Max,

I'm the author of this post:
http://blog.notdot.net/2009/11/Damn-Cool-Algorithms-Spatial-indexing-with-Quadtrees-and-Hilbert-Curves

You're right that this will likely have problems when dealing with arbitrary
data rather than geospatial indexing. Hilbert curves work well at returning
results with a minimum of false-positives and minimal queries when the
region you're searching is (in coordinates on the curve) roughly square, or
at least no more than 2:1. With arbitrary indexes, you could easily have
queries that don't fit that approximation. At the very least, it would
require some tuning.

-Nick Johnson

On Tue, Jul 5, 2011 at 12:22 PM, Max <[email protected]> wrote:

> Hi all,
>
> Would like to know if there are any of you guys ever tried to use space
> filling curve like Hilbert curve to build index for multiple inequality
> filters.
>
> Seems like for any continuous field like long or date, the number of ranges
> (to be merged) to perform a accurate query is increasing rapidly, which
> makes this approach not scale.
>
> Any thought? or shall I build / model like this 
> sample<http://code.google.com/appengine/articles/geosearch.html> app?
> Query by partitions of all data and do an in-memory merge?
>
> --
> You received this message because you are subscribed to the Google Groups
> "Google App Engine" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/google-appengine/-/MipYkQ1_pvYJ.
> 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/google-appengine?hl=en.
>



-- 
Nick Johnson, Developer Programs Engineer, App Engine

-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" 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/google-appengine?hl=en.

Reply via email to