Skip to main content
Geographic Information Systems

Return to Answer

deleted 426 characters in body
Source Link
user1702401
user1702401
  • 2.9k
  • 1
  • 19
  • 18

You could try canvas renderer, that improves rendering speed (but has several other disadvantages, like problems with selecting feature etc).

Also, usually so big datasets are displayed through WMS service (rendering is done on server side), are there any particular reasons, why you can't do that?

To reduce point count on map, you can use clustering strategy (http://dev.openlayers.org/releases/OpenLayers-2.12/examples/strategy-cluster.html). Since 14000 features is quite a lot, it may still be slow because of clustering computations.

Also, you can write some point reduction code yourself. For example: store all features in array. Figure out, how you want to reduce features - by some attributes, by viewport boundary, etc. On certain map events (like moveend, when you are restricting by viewport; on zoomend, when you are doing kind of zoomlevel-based clustering) remove all features from layer and add reduced set of new ones, that are read from array.

You could try canvas renderer, that improves rendering speed (but has several other disadvantages, like problems with selecting feature etc).

Also, usually so big datasets are displayed through WMS service (rendering is done on server side), are there any particular reasons, why you can't do that?

To reduce point count on map, you can use clustering strategy (http://dev.openlayers.org/releases/OpenLayers-2.12/examples/strategy-cluster.html). Since 14000 features is quite a lot, it may still be slow because of clustering computations.

Also, you can write some point reduction code yourself. For example: store all features in array. Figure out, how you want to reduce features - by some attributes, by viewport boundary, etc. On certain map events (like moveend, when you are restricting by viewport; on zoomend, when you are doing kind of zoomlevel-based clustering) remove all features from layer and add reduced set of new ones, that are read from array.

You could try canvas renderer, that improves rendering speed (but has several other disadvantages, like problems with selecting feature etc).

Also, usually so big datasets are displayed through WMS service (rendering is done on server side), are there any particular reasons, why you can't do that?

To reduce point count on map, you can use clustering strategy (http://dev.openlayers.org/releases/OpenLayers-2.12/examples/strategy-cluster.html). Since 14000 features is quite a lot, it may still be slow because of clustering computations.

added 146 characters in body
Source Link
user1702401
user1702401
  • 2.9k
  • 1
  • 19
  • 18

You could try canvas renderer, that improves rendering speed (but has several other disadvantages, like problems with selecting feature etc).

Also, usually so big datasets are displayed through WMS service (rendering is done on server side), are there any particular reasons, why you can't do that?

To reduce point count on map, you can use clustering strategy (http://dev.openlayers.org/releases/OpenLayers-2.12/examples/strategy-cluster.html). Since 14000 features is quite a lot, it may still be slow because of clustering computations.

Also, you can write some point reduction code yourself. To give some ideas, clustering strategy works soFor example: store all features are stored in array array. Figure out, how you want to reduce features - by some attributes, by viewport boundary, etc. On eachcertain map zoom eventevents (like moveend, featureswhen you are restricting by viewport; on layer are removedzoomend, when you are doing kind of zoomlevel-based clustering) remove all features from layer and add reduced set of new ones are added, based on calculations to reduce count of them,that are read from array.

You could try canvas renderer, that improves rendering speed (but has several other disadvantages, like problems with selecting feature etc).

Also, usually so big datasets are displayed through WMS service (rendering is done on server side), are there any particular reasons, why you can't do that?

To reduce point count on map, you can use clustering strategy (http://dev.openlayers.org/releases/OpenLayers-2.12/examples/strategy-cluster.html). Since 14000 features is quite a lot, it may still be slow because of clustering computations.

Also, you can write some point reduction code yourself. To give some ideas, clustering strategy works so: all features are stored in array. On each map zoom event, features on layer are removed and new ones are added, based on calculations to reduce count of them, from array.

You could try canvas renderer, that improves rendering speed (but has several other disadvantages, like problems with selecting feature etc).

Also, usually so big datasets are displayed through WMS service (rendering is done on server side), are there any particular reasons, why you can't do that?

To reduce point count on map, you can use clustering strategy (http://dev.openlayers.org/releases/OpenLayers-2.12/examples/strategy-cluster.html). Since 14000 features is quite a lot, it may still be slow because of clustering computations.

Also, you can write some point reduction code yourself. For example: store all features in array. Figure out, how you want to reduce features - by some attributes, by viewport boundary, etc. On certain map events (like moveend, when you are restricting by viewport; on zoomend, when you are doing kind of zoomlevel-based clustering) remove all features from layer and add reduced set of new ones, that are read from array.

Source Link
user1702401
user1702401
  • 2.9k
  • 1
  • 19
  • 18

You could try canvas renderer, that improves rendering speed (but has several other disadvantages, like problems with selecting feature etc).

Also, usually so big datasets are displayed through WMS service (rendering is done on server side), are there any particular reasons, why you can't do that?

To reduce point count on map, you can use clustering strategy (http://dev.openlayers.org/releases/OpenLayers-2.12/examples/strategy-cluster.html). Since 14000 features is quite a lot, it may still be slow because of clustering computations.

Also, you can write some point reduction code yourself. To give some ideas, clustering strategy works so: all features are stored in array. On each map zoom event, features on layer are removed and new ones are added, based on calculations to reduce count of them, from array.