A pathfinding algorithms visualizer
This application was developed with pygame, is used to visualizer all of pathfinding algorithms (BFS, DFS, A*, etc).
In this program, each grid on map can be expanded in 8 directions (For 4 straight directions, switch to "no-diagonal" branch). The purpose is to see the difference between difference informed and uninformed pathfinding algorithms, when each move can has different costs, and shortest path is not necessarily the path that has least moves.
How to use the program:
- Install necessary packages:
pip install -r requirements.txt - Run the simulation: python game.py
Some notes for the program:
-
When starting, set up starting and goal nodes (using mouse). Then, set barricades/walls (using mouse). Then, hit Enter to solve it. Press Escape to escape the program.
-
Legends:
- Red node: Starting node
- Green node: Goal node
- Gray nodes: Barricades/Walls
- Yellow nodes: Explored nodes
- Blue nodes: Frontier (Potential) nodes
- Purple nodes: The path that was found by the algorithm.
Testing result on one example input for different algorithms:
- BFS (Breadth-First Search):
Cost: 2711 px
Iteration: 1798
- Greedy Best-First Search (f(n) = h(n)):
Cost: 2259 px
Iteration: 83
- A* (f(n) = g(n) + h(n)):
Cost: 2125 px
Iteration: 530
- Uniform-cost search (f(n) = g(n)):
Cost: 2090 px
Iteration: 1835
Note:
- Iteration here is the number of iteration through frontier (reachable nodes from any node) list through expansion, before the algorithm reachs the goal.
- The heuristic function (h(n)) is used in this problem is Euclidean distance between the node and goal node, since the distance of one node to another could be different.
- Path cost (g(n)) of a node is calculated by the sums of Euclidean distance between one node and parent node (Do it from the current node, traverse back until there is no more parent node)