A fast and memory-efficient 2D RTree implementation for Javascript. Supports range searches and nearest neighbour searches.
The RTree is available as:
- script: ./release/js/RTree2d.js (adds a global variable with the name
RTree2d) - an AMD module: ./release/js/RTree2d-amd.js
- an CommonJS module: ./release/js/RTree2d-common.js
Take a test drive here.
var rtree = new RTree2d();
//insert an object at the rectangle (x/y/width/height).
var ob = {store: 'me'};
rtree.insert(ob,2,3,4,5);
//search a rectangle
var results = rtree.search(0,0,10,10);
//results ~= [{store: "me"}];
//Instead of using search, you can also use one of the collection comprehension methods.
//(e.g. you might be doing a search 60 frames-per-second. You do not want a new array each time).
rtree.forEachInRectangle(0,0,10,10, function(ob){
console.log("i got a hit, ",ob);
});
//find the k nearest neighbours (x/y/k)
var kNNArray = rtree.nearestNeighbours(0,0,1);
//Number of elements in the tree.
var size = rtree.size();
//Remove an element.
rtree.remove(ob,2,3,4,5);
-
checkout repo
-
run node setup script (might need admin priviliges on windows)
node setup.js
-
build a release
grunt release