• Basic grid-based sandbox
• Options to choose between the different search implementations
• Some text and documentation
• Available independently on gh-pages

Necessary to allow step by step scenarios

Have you considered switching out from your own priority queue implementation to a Fibonacci Heap implementation? FH has one of the best amortized runtimes for Dijkstra implementations.

The mnemonist library has an implementation that's stable.

• A* (by just adding a simple heuristic function parameter to Dijkstra?)
• Bidirectionnal Dijkstra (and A*)
• Dijkstra on edges (a la PG Routing Shooting Star)
• Choice routing?

Example:

• matrix based
• shared adjacency list instead of lists by node
• efficient storage for undirected graphs

Modules and NPM and stuff :)