[
https://issues.apache.org/jira/browse/LUCENE-6699?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14644248#comment-14644248
]
Karl Wright commented on LUCENE-6699:
-------------------------------------
bq. Can we "just" index earth surface lat/lon, and then at query time is
spatial3d able to
give me an enclosing "surface lat/lon" bbox for a 3d shape?
Ok, so now you have me confused a bit as to what your requirements are for BKD.
If you want to split the globe up into lat/lon rectangles, and use BKD that
way to descend, then obviously you'd need points to be stored in lat/lon. But
that would make less sense for geospatial3d, because what you're really trying
to do is assess membership in a shape, or distance also in regards to a shape,
both of which require (x,y,z) not lat/lon. Yes, you can convert to (x,y,z)
from lat/lon, but the conversion is relatively expensive.
Instead, I could imagine just staying natively in (x,y,z), and doing your
splits in that space, e.g. split in x, then in y, then in z. So you'd have a
GeoPoint3D which would pack (x,y,z) in a format you could rapidly extract, and
a splitting algorithm that would use the known ranges for these values. Does
that make sense to you? Would that work with BKD?
> Integrate lat/lon BKD and spatial3d
> -----------------------------------
>
> Key: LUCENE-6699
> URL: https://issues.apache.org/jira/browse/LUCENE-6699
> Project: Lucene - Core
> Issue Type: New Feature
> Reporter: Michael McCandless
> Assignee: Michael McCandless
>
> I'm opening this for discussion, because I'm not yet sure how to do
> this integration, because of my ignorance about spatial in general and
> spatial3d in particular :)
> Our BKD tree impl is very fast at doing lat/lon shape intersection
> (bbox, polygon, soon distance: LUCENE-6698) against previously indexed
> points.
> I think to integrate with spatial3d, we would first need to record
> lat/lon/z into doc values. Somewhere I saw discussion about how we
> could stuff all 3 into a single long value with acceptable precision
> loss? Or, we could use BinaryDocValues? We need all 3 dims available
> to do the fast per-hit query time filtering.
> But, second: what do we index into the BKD tree? Can we "just" index
> earth surface lat/lon, and then at query time is spatial3d able to
> give me an enclosing "surface lat/lon" bbox for a 3d shape? Or
> ... must we index all 3 dimensions into the BKD tree (seems like this
> could be somewhat wasteful)?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
