If you are interested in using MapServer, it now supports clustering of points.
http://mapserver.org/mapfile/cluster.html David. On Tue, Sep 6, 2011 at 8:26 AM, Michael <[email protected]> wrote: > The thing is that 20,000 points is more points than anyone can reasonably > perceive on the screen to get any useful information. You are on the right > track to consider clustering. > > My application (tracking 70 boats across the Pacific) gives me a raw data > set of about 15,000 points each with lat, lon, time, and boat id. That's a > lot of data to process and to upload. > > What I am doing instead is pre-thinning the data using the Douglas-Peuker > algorithm on the server side, which nicely reduces the data to any level of > detail you care to specify. I got about a 90% reduction in point count with > no real loss of information on tracks. > > If your data is not lines, but discrete points, then I am less versed. > Yeah, clustering sounds good. You may also want to use the BBOX strategy to > only fetch the points that are relevant to your map. > > I can post the douglas-peuker code somewhere if that would be of value. > > Michael > > On 9/6/2011 6:15 AM, Tito, Joseph wrote: > > I’m trying to expand one of my current application to handle larger > amounts of data. Right now the application comfortably handles 2000-3000 > points at a time, but I'd like to expand its capabilities to handle at LEAST > 20,000 points. More would be nice, but the minimum requirement is 20,000. > Displaying these 20,000 points as X number of ‘clusters’ is totally > acceptable, as long as it gives a good visual representation of the full > data set. > > > > One thing I need to take into consideration is that I have multiple data > types represented as their own layers, and each user has the ability to turn > on/off layers based on their personal preferences. > > My question to everyone is, how have others solved this problem? I have a > couple server side ideas to "pre-cluster" data in mind, but I want to see > how others have tackled this problem before I attack it. If anyone has links > to other relevant posts or websites, please share! > > > > Thank you > > > > Joe Tito > > _______________________________________________ > Users mailing list > [email protected] > http://lists.osgeo.org/mailman/listinfo/openlayers-users > > _______________________________________________ > Users mailing list > [email protected] > http://lists.osgeo.org/mailman/listinfo/openlayers-users > > _______________________________________________ Users mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/openlayers-users
