@robertaron/spacial-partitioning

Efficient spatial partitioning and neighbor graph generation for 3D simulations, powered by Rust and WebAssembly. This package provides a fast way to compute nearby relationships between points in 3D space, useful for flocking, boids, crowd simulation, and other spatial algorithms.

Features

• High performance: Core logic implemented in Rust and compiled to WebAssembly.
• Works in Node.js and browsers: Supports both ESM and CommonJS.
• Typed API: TypeScript types included.

Installation

npm install @robertaron/spacial-partitioning

Usage

import { init, createNearByGraph } from "@robertaron/spacial-partitioning";

// Initialize the WASM module (required before using createNearByGraph)
await init();
const positions = new Float32Array([
  0, 0, 0,
  1, 0, 0,
  0, 10, 0,
  0, 10, 1,
  // ... more points
]);
const neighborGraph = createNearByGraph(positions, 2);
// neighborGraph is an array of arrays: neighborGraph[i] = [indices of neighbors of i]

API

async init(): Promise<void>

Initializes the WASM module. Must be called before using createNearByGraph.

createNearByGraph(positions: Float32Array, distance: number): number[][]

• positions: Flat array of 3D coordinates (x, y, z, x, y, z, ...).
• distance: Maximum distance to consider two points as neighbors.
• Returns: An array of arrays, where each sub-array contains the indices of neighboring points.

Use Cases

• Flocking/boids simulations
• Crowd and agent-based modeling
• Efficient spatial queries in 3D environments

License

MIT

Credits

• Rust WASM core by @robertaron
• Inspired by spatial partitioning techniques in game development and simulation

For more details,