I would like to query my latitude/longitude points for distance in a
way that uses the index. The points are in lat/lon, so st_dwithin(g1,
g2, [meters]) makes no sense.
What is the recommended way to do this? I think what I need is a
function that takes a point in lat/lon, and a distance in meter
http://trac.osgeo.org/postgis/wiki/UsersWikiExamplesFindNearbyLatLon
On Wed, Jul 22, 2009 at 11:19 PM, Jeff Davis wrote:
> I would like to query my latitude/longitude points for distance in a
> way that uses the index. The points are in lat/lon, so st_dwithin(g1,
> g2, [meters]) makes no sense.
>
On Wed, 2009-07-22 at 23:37 -0700, Paul Ramsey wrote:
> http://trac.osgeo.org/postgis/wiki/UsersWikiExamplesFindNearbyLatLon
>
Ok, that looks like it will work.
It seems like there should be some calculation that gives a tight
bounding box for any point, is there a reason that's hard to find?
R
Jeff,
* Jeff Davis (pg...@j-davis.com) wrote:
> I would like to query my latitude/longitude points for distance in a
> way that uses the index. The points are in lat/lon, so st_dwithin(g1,
> g2, [meters]) makes no sense.
>
> What is the recommended way to do this? I think what I need is a
> func
On Thu, 2009-07-23 at 07:06 -0400, Stephen Frost wrote:
> Why do you care about it being in meters..?
Well, all I care about is that it's in some kind of linear length units.
Queries might take the form of "what else is within 3 miles of this
lat/lon point?".
> Just make sure that
> st_expand wi
Just isn't such a thing, is all. I guess an st_expand_sphere() would
be what you wanted, but does not exist. Hopefully GEOGRAPHY will
retire all these requirements once and for all :)
P
On Thu, Jul 23, 2009 at 8:49 AM, Jeff Davis wrote:
> On Thu, 2009-07-23 at 07:06 -0400, Stephen Frost wrote:
>>
On Thu, 2009-07-23 at 08:58 -0700, Paul Ramsey wrote:
> Hopefully GEOGRAPHY will retire all these requirements once and for
> all :)
Is that something under development? I'm new to these lists, so I
haven't been following.
Regards,
Jeff Davis
_
I'm facing a similar issue; a huge postgis database with a point
column, trying to find a clean query to bring back the nearest n
results to a point. Why isn't that easy? Or is it?
I'm using georuby and spatial adapter.
Steve
On Thu, Jul 23, 2009 at 10:01 AM, Jeff Davis wrote:
> On Thu, 2009-07
steve enzer wrote:
I'm facing a similar issue; a huge postgis database with a point
column, trying to find a clean query to bring back the nearest n
results to a point. Why isn't that easy? Or is it?
The problem is that the nearest n points is not well defined in spatial
terms. For example th
Nearest neighbor, as Stephen notes in the followup is not something
that is necessarily easy in general. However! Another development
on-stream for PostgreSQL 8.5 will allow us to both (a) use k-d tree
indexes for points, which are much more efficient and (b) to do
nearest-neighbor searching using
Yes, under development, with a client delivery date in November and
hopefully rolled into a new numbered release (postgis 1.5) before the
new year.
P.
On Thu, Jul 23, 2009 at 10:01 AM, Jeff Davis wrote:
> On Thu, 2009-07-23 at 08:58 -0700, Paul Ramsey wrote:
>> Hopefully GEOGRAPHY will retire all
Oh, and the basic sql query would be:
select * from poi_table where st_within(setsrid(makepoint(x,y),4326),
the_geom, radius) order by st_distance(setsrid( makepoint(x,y),4326),
the_geom) asc limit n;
(x,y) == (lon, lat)
radius == is your initial search radius
n == is your limit on results
-
Whooops, you are right. Change that query to be:
select * from poi_table
where
st_within(the_geom, st_expand(setsrid(makepoint(x,y),4326), radius))
order by st_distance(setsrid( makepoint(x,y),4326), the_geom) asc
limit n;
or to explicitly specify the index usage
select * from poi_table
13 matches
Mail list logo
