Algorithmica

An interactive educational sandbox for demonstrating and experimenting with procedural map generation algorithms.

1. Binary Space Partitioning — Dungeon Generation

The algorithm recursively divides the map into rectangular sections. Each leaf of the split tree becomes a room, and corridors are carved between adjacent rooms. The depth of the recursion determines the number of rooms and the overall complexity of the maze.

2. Cellular Automata — Landscape Generation

Starting from a random noise pattern (land vs. water), the algorithm applies several smoothing passes. Isolated cells disappear, large connected areas grow, and the chaos gradually turns into coherent terrain — islands, caves, peninsulas. Post-processing ensures every landmass is reachable.

3. Wave Function Collapse — Advanced Dungeon Generation

Wave Function Collapse is a constraint-solving algorithm inspired by quantum mechanics. Every cell on the grid starts in a "superposition" of all possible states. As cells collapse one by one, the algorithm propagates constraints to their neighbors, resolving the grid into a locally consistent layout — though a plain WFC run offers no guarantee of a solvable, fully connected structure: contradictions can occur, forcing a backtrack or a full regenerate.

To solve this, a hybrid topology layer is built on top of standard WFC; in this project, a BSP tree is converted into a level topology graph, and key paths are pre-constrained ahead of the collapse — guaranteeing dungeon connectivity without relying on backtracking.

Interactive Features

• Parameter Tooltips: Hover over any config field to see what it actually does — no guesswork.
• Diagnostics & Performance: Keep track of generation time (ms) and RAM usage for each layout.
• Deterministic Seeds: Use seed values to recreate identical layouts — perfect for testing or sharing.
• Real-Time Tweaking: Adjust map parameters on the fly and watch the map rebuild instantly.

For Developers & Open Source

Algorithmica is fully open-source, written in clean, modular C#. Use it as an interactive reference for integrating procedural generation algorithms into Godot 4+ projects.

Repository & Code: https://github.com/Holizer/procgen-lab