This module allows one to execute A* search in an Asynchronous pattern. It is very similar to EasyStar, however this module can be used for more than 2D planning for game development as EasyStar is designed for. It allows one to specify a generic Generate Successors function, not needing to know anything about the graph. In other words, this module only cares about the algorithm of A*, not the graph, which is from your specific problem domain. It could be 2D, 3D, or however many dimensions you may need (though A* performance suffers greatly at higher dimensions).

Install : npm install asyncastar

Since most people will probably do 2D or 3D path planning, there is a simple wrapper for AsyncAStar called createPlanner that assumes a 3D discretized space using ndarray as the data holder of this space. The demonstration below shows 3D planning within a plane (basically 2D).

import { createPlanner} from '../lib/util';
const start = [0, 0, 0]
const goal = [2, 2, 0]
const maze = ndarray([0,0,0,0,0,0,0,0,0], [3,3,1]) // Empty 3X3X1 Maze
planner = createPlanner(maze, mazeTest.start, mazeTest.goal);

let result = planner.searchAsync(1) // only 1 expansion
console.log(result)


result = planner.searchAsync(3) // allow 3 more expansions
console.log(result)

Full API: https://jeremybyu.github.io/AsyncAstar/

Feel free to file any issues (bugs), or feature requests. Please follow the Issue template however.

First download, pull, and run tests. Make sure its all working!

1. git pull https://github.com/JeremyBYU/AsyncAstar.git
2. npm install
3. npm run test

Then submit any PR's, following the contributing guidelines.

X and Y might be flipped in NDARRAY