WFC

Overlapping Wave Function Collapse algorithm implemented in Elixir.
One of the main goal is readability so codebase can be used for learning the algorithm.

Paper to help understand the algorithm

Wave Function Collapse repo

Try it out!

This will generate an example png for you in main dir.

Wfc.example()

How it works?

Let’s say you have an input like the below one:

  [
    [{255, 255, 255}, {255, 255, 255}, {255, 255, 255}, {255, 255, 255}],
    [{255, 255, 255}, {0, 0, 0}, {0, 0, 0}, {0, 0, 0}],
    [{255, 255, 255}, {0, 0, 0}, {255, 0, 0}, {0, 0, 0}],
    [{255, 255, 255}, {0, 0, 0}, {0, 0, 0}, {0, 0, 0}]
  ]

First you create NxN tiles from it. (optionally to make more tiles you rotate them around)
This happens in Wfc.Input.tiles/2.

[
  [[{255, 255, 255}, {255, 255, 255}], [{255, 255, 255}, {0, 0, 0}]],
  [[{255, 255, 255}, {255, 255, 255}], [{0, 0, 0}, {0, 0, 0}]],
  [[{255, 255, 255}, {0, 0, 0}], [{255, 255, 255}, {0, 0, 0}]],
  [[{0, 0, 0}, {0, 0, 0}], [{0, 0, 0}, {255, 0, 0}]],
  ...
]

Now you need to determine which tiles can go next to each other in which directions.
This happens in Wfc.Input.constraints/2.
For example if you want to determine that b_tile can go above a_tile:

  [a_tile_top_row | _] = a_tile = [[{255, 255, 255}, {255, 255, 255}], [{255, 255, 255}, {0, 0, 0}]],
  [_top_row | b_tile_bottom_row] = b_tile = [[{255, 255, 255}, {255, 255, 255}], [{255, 255, 255}, {0, 0, 0}]],
  can_b_tile_go_on_top_of_a_tile? = (a_tile_top_row == b_tile_bottom_row)

Now that you have your tiles with your constraints:

1. You create your output matrix and you set all elements to include all possible tiles. Wfc.Output.new/2
2. You set a random element to just 1 possible tile, and you propagate the changes to the neighbours. Wfc.Output.collapse/1
3. When the above propagation is ready, you select the element with lowest possibilities and you repeat step 2 until you have only elements with 1 possible tile. Wfc.Output.lowest_possibilities_position/1
4. You are done.