pts-to-graph

generate nearest neighbors for a set of 3d points.

convert triangle mesh to graph.

convert set of 3d points into an ngraph.

Installation

npm i pts-to-graph

Usage - list of points

var listOfPts = [];

var nPts = 1000;
var size = 1024;

for(var i=0; i<nPts; i++){
    var pt = [Math.random()*size, Math.random()*size, Math.random()*size]
    listOfPts.push(pt);
}

var minimumDist = 200; //can either set minimumDist, or leave out the parameter and set pt[i].scale=distance to have each pt have a different neighbor radius
p2g.setOfPtsNeighborsOctree(listOfPts, minimumDist); //fast - get neighbors using octree data structure [npm yaot] - 20ms
//p2g.setOfPtsNeighborsRTree(listOfPts, minimumDist); //fast - get neighbors using rtree data structure [npm rbush-3d] - 31 ms
//p2g.setOfPtsNeighborsDelaunay(listOfPts); //very slow - get neighbors using delaunay triangulation - 2144 ms on m1 mac

//var myNGraph = p2g.setOfPtsToNGraph(listOfPts); //convert points with neighbors into ngraph graph

//results
//listOfPts[0].neighbors -- array of points nearby [by reference]
//listOfPts.octree -- octree from require('yaot')
//listOfPts.rtree -- rtree from require('rbush-3d')

Usage - triangle mesh

var p2g = require('pts-to-graph');
var mesh = require('bunny');
var sop = p2g.mesh2PtsWithNeighbors(mesh);
console.log(sop);

//conver to ngraph with p2g.setOfPtsToNGraph

//      [
//         [ 1.301895, 0.122622, 2.550061, neighbors: [Array], i: '0' ],
//         [ 1.045326, 0.139058, 2.835156, neighbors: [Array], i: '1' ],
//         [ 0.569251, 0.155925, 2.805125, neighbors: [Array], i: '2' ],
//         ... ]

Usage - set of line segments

var p2g = require('pts-to-graph');

//a set of 3d pts...
var a = [0,0,0];
var b = [0,0,1];
var c = [0,1,1];
var d = [0,1,0];

//a set of 3d line segments...
var lines = [
    [a,b],
    [b,c],
    [c,d],
    [d,a]
]

// note - also works with raw coords - points get de-duplicated
// var lines = [
//     [[0,0,0],[0,0,1]],
//     [[0,0,1],[0,1,1]],
//     [[0,1,1],[0,1,0]],
//     [[0,1,0],[0,0,0]]
// ]

//linesToSetOfPtsNeighbors(lines, eps=0.01, deCleanup=true) 
//      - eps used for point de-duplicating and rtree radius padding
//      - if doCleanup=false, pt[i] will contain .origPt and .ptIndex [used internally by algorithm]
var ptsWithNeighbors = p2g.linesToSetOfPtsNeighbors(lines);
console.log(ptsWithNeighbors);
// [
//     [ 0, 0, 0, neighbors: [ [Array], [Array] ], i: '0' ],
//     [ 0, 0, 1, neighbors: [ [Array], [Array] ], i: '1' ],
//     [ 0, 1, 1, neighbors: [ [Array], [Array] ], i: '2' ],
//     [ 0, 1, 0, neighbors: [ [Array], [Array] ], i: '3' ]
// ]


// >> you can also do the reverse - convert points-with-neighbors back into list of line segments

var lines2 = p2g.setOfPtsNeighborsToLines(ptsWithNeighbors);

console.log(lines2);
// a list of line segments, made from the same points [note how they still have neighbors added]
// [
//     [
//         [ 0, 0, 0, neighbors: [Array], i: '0' ],
//         [ 0, 0, 1, neighbors: [Array], i: '1' ]
//     ],
//     [...

// >> convert a mesh into a set of lines [triangle edges become lines, eg if we want to disgard some edges before converting into a graph]

var mesh = require('bunny');
var lines3 = p2g.meshToLines(mesh);

See Also

• yaot
• rbush-3d
• ngraph