ajnirp / maze-generation

Implementations of maze generation algorithms, inspired by Jamis Buck's blogposts

About

Implementations of maze generation algorithms. Some are just direct Python ports of Jamis Buck's implementations, for example the recursive backtracking algo.

Demo

Cartesian maze

Triangle maze

Hex maze

Setup

virtualenv .
. bin/activate
pip install Pillow
mkdir img

Run

python main.py

TODO

• Add code to print the maze parameters at the bottom of the generated image, with a link to my Github?
• Factor out maze generation methods into their own class structure?
• Command-line args to specify what kind of maze and what maze generation method and what maze size
  • Check for compatibility: some mazes might permit only certain methods?
  • Ideally, user specifies maze type, size and window dimensions, and we calculate suitable values for scale and padding from those params
• Improve rendering: fill a non-white background color for the maze?
• Add demo images to the README

More long-term ideas

• Generate a maze and send it to an algorithm that does path-finding
• Implement different path finding algos like DFS, BFS, A* maybe?
• Visualize the found paths and maybe animate it with the turtle library?
• Smooth mazes with arbitrary degree turns? Maybe generate a rectangular maze and then smoothen it somehow? Circular maze? Crude way to model a side cartoon view of the human brain?