locked
Multimap Clustering RRS feed

  • Question

  • The default clustering options for decluttering markers seems to work very well for any zoom level above 7 but not below (bing maps seems to work well on all zoom levels)

    The scenario is that we have 400 markers being plotted on a map of the UK and we obviously need to cluster these up at certain zoom levels. The default clustering options seem to work very well for any zoom level above 7 but not below.

    There is a caviate to this though when I'm on zoom level 8 (around 50 clusters) and clustering is working well and I zoom out to zoom level 7 then the clustering appears to cluster correctly (more like bing) for a split second (around 20 clusters) and then seems to cluster again and the results of the clustering end up with 4 clustered markers one containing 268 markers another containing 4, another containing 34 and another containing 31.

    Have you come across this issue before and how was it handled?
    Wednesday, September 23, 2009 3:17 PM

Answers

  • The default clustering options for decluttering markers seems to work very well for any zoom level above 7 but not below (bing maps seems to work well on all zoom levels)

    The scenario is that we have 400 markers being plotted on a map of the UK and we obviously need to cluster these up at certain zoom levels. The default clustering options seem to work very well for any zoom level above 7 but not below.

    There is a caviate to this though when I'm on zoom level 8 (around 50 clusters) and clustering is working well and I zoom out to zoom level 7 then the clustering appears to cluster correctly (more like bing) for a split second (around 20 clusters) and then seems to cluster again and the results of the clustering end up with 4 clustered markers one containing 268 markers another containing 4, another containing 34 and another containing 31.

    Have you come across this issue before and how was it handled?
    Hi Stav
    You are experiencing the expected behaviour of the clustering algorithm.
    The clustering algorithm works by finding and creating a virtual boundary around a marker which is slightly bigger than the marker itself. If any other markers are found within that boundary then they are added to the cluster and the boundary of the cluster is extended to cover an area slightly bigger than the original positions of all the markers in the cluster. If lots of markers occupy a relatively small area of the map and are close to each other then they’ll end up being put into the same cluster.

    We recommend trying the different options available on the clustering call, maybe try using the grid clustering method or specifying accurate clustering which might improve the results.
    Further information is in the class docs:
    http://clients.multimap.com/share/documentation/api/1.2/classes/MultimapViewer.html#declutterGroup



    Thursday, September 24, 2009 1:38 PM
    Moderator