JSPM

ngraph.quadtreebh3d

1.0.0
  • ESM via JSPM
  • ES Module Entrypoint
  • Export Map
  • Keywords
  • License
  • Repository URL
  • TypeScript Types
  • README
  • Created
  • Published
  • Downloads 1382
  • Score
    100M100P100Q100538F
  • License MIT

Quad Tree data structure for Barnes-Hut simulation in 3d space

Package Exports

  • ngraph.quadtreebh3d

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (ngraph.quadtreebh3d) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

ngraph.quadtreebh3d

Quad Tree data structure for Barnes-Hut simulation in 3d space. Technically it should be named OctTree, since it's not a quad tree. But for parity with 2d interface it's currently called quadtree. Let me know if you think it should be changed :).

This project is part of ngraph family. If you need 2d quad tree follow to ngraph.quadtreebh.

Build Status.

Reference

  1. Fast Hierarchical Methods for the N-body Problem - one of the best explanations of Barnes-Hut method by James Demmel.
  2. Wikipedia article - general introduction into the problem

Usage

var Body = require('ngraph.physics.primitives').Body3d;
// Create new bodies at (1, 1, 1) and (0, 0, 0):
var bodies = [];
bodies.push(new Body(1, 1, 1));
bodies.push(new Body(0, 0, 0));
/* ... create more as you need ... */

// build quad tree:
var createQuadTree = require('ngraph.quadtreebh3d');
var quadTree = createQuadTree();

// insert bodies into the quad tree 
quadTree.insertBodies(bodies); // performance: O(n * log n)

// calculate forces acting on each body in the tree O(n * log n):
bodies.forEach(function(body) {
  quadTree.updateBodyForce(body);
});
// At this point every body object has valid 3d force vecor
console.dir(bodies[0].force);

Note: You don't necessary have to use ngraph.physics.primitives. That package merely defines an interface for a physical Body, which is expected by quad tree. As long as your Body object implements this interface (mass, pos and force) - you can use current quad tree structure to calculate forces.

Configuring quad tree

Quad tree allows to change two global options:

  • gravity - Gravitational constant, used in force value calculation. Defaults to -1;
  • theta - a threshold which determines when group of bodies is considered distant enough to be approximated as a single body. The value should be > 0, and usually little less than 1. It defaults to 0.8 if not suplied. This argument affects accuracy and speed of simulation. Please refer to [1] for more details.

You can pass these setting to quad tree as follows:

var createQuadTree = require('ngraph.quadtreebh3d');
var quadTree = createQuadTree({
  theta: 1.2,
  gravity: -10
});

To query current options of the tree use:

var createQuadTree = require('ngraph.quadtreebh3d');
var quadTree = createQuadTree();
console.dir(quadTree.options()); // prints { theta: 0.8, gravity: -1};

To change options at run time:

var createQuadTree = require('ngraph.quadtreebh3d');
var quadTree = createQuadTree();
quadTree.options({
  theta: 0.5,
  gravity: -42
});

Mass of each body affects overall result of computation. You can tweak it when creating new bodies:

var Body = require('ngraph.physics.primitives').Body;
var earth = new Body(); earth.mass = 5.972;
var sun   = new Body(); sun.mass = 1989000;

install

With npm do:

npm install ngraph.quadtreebh3d

license

MIT