Graph Search Toy

A toy I built to demonstrate how different graph search functions work, and the strengths and weaknesses of each of them.

Includes: Breadth-First Search, Depth-First Search, Uniform-Cost Search, Heuristic Search, and A* Search, with loads of parameters to customize the specifics.

It also contains several heuristics to choose from, some admissible, some inadmissible. A good example of an inadmissible heuristic failing is using A* with Double Manhattan on the 6th level.

All edges that go into a node that is at a slime tile have a weight of 2, supposedly because the slime tiles make things move twice as slow. All edges that go into a node that is at a fire tile have a weight of 20, supposedly because going into a fire is something we should try to avoid if we can, but not to an absurd degree. (So it's like 1HP = 20 steps.)

This program was written for a presentation I gave as a part of a lab section on A* for my job as a teacher's assistant for the Game Development course at IU.

The "story" of the game is that it's a standard sokoban game, but there is no player, so the box has to become sentient and push itself to the target.