pathfinding-visualizer

Grid-based visualizer for popular pathfinding algorithms. Written in Python, with PyQt5 for rendering and UI.

Left-click on empty tiles to begin drawing barriers, left-click on barriers to begin erasing them. Left-click on start or goal to drag them around. Select different algorithms and heuristics using the menubar.

Implementation

Algorithms

• A* Search
  • Searches nodes based on f(n) = cost_so_far(n) + heuristic(n)
  • Weighted, always optimal given an admissible heurisitic
• Dijkstra's algorithm
  • Searches nodes by progressively moving further from the starting node
  • Weighted, always optimal
• Greedy Best-First Search
  • At every step, explore the best node according to chosen heuristic
  • Weighted, not always optimal
• Breadth-First Search
  • Nodes are progressively explored by adding neighbors to a queue
  • Unweighted, always optimal given an unweighted graph
• Depth-First Search
  • Nodes are progressively explored by pushing neighbors to a stack
  • Unweighted, not always optimal

Heuristics

• Manhattan: distance between two nodes measured along axes
• Euclidean: straight-line distance between two nodes
• Chebyshev: distance between two nodes measured along axes, but diagonal movement costs 1
• Octile: distance between two nodes measured along axes, diagonal movement costs √2

Usage

$ cd <path of project install> $ python3 pathfinding-visualizer.py

Upcoming Features

• Support for visualizing and modifying individual cell weights
• Legend for different node colors
• In-app display of each algorithm's characteristics
• Create your own heuristic formulas
• More algorithm options

About

License: This project is released under the MIT License.

References:

• Amit's A* Pages