bi3mer / GramElites

Gram-Elites

A free to read version of the paper can be found here. If you want to use generated segments without running the code yourself, you can get the data here,

Abstract

In the context of procedural content generation via machine learning (PCGML), quality-diversity (QD) algorithms are a powerful tool to generate diverse game content. A branch of QD uses genetic operators to generate content (e.g. MAP-Elites). Problematically, levels generated with these operators have no guarantee of matching the style of a game. This can be addressed by incorporating whether a level is generable by an n-gram into the fitness function. Unfortunately, this leads to wasted computation and poor results. In this work, we introduce n-gram genetic operators, which produce only solutions that are generable by the n-gram model; we call MAP-Elites combined with these operators Gram-Elites. We test on a tile-based side-scrolling platformer, vertical platformer, and roguelike. For all three, n-gram operators outperform standard operators and random n-gram generation, finding more usable (i.e. completable and generable) solutions at a faster rate. By integrating structure into operators, instead of fitness, these genetic operators could be beneficial to QD in PCGML.

Use

GAME can be "mario", "icarus", or "dungeongram". ALGORITHM can be "n-gram-placement", "map-elites", or "gram-elites".

Generate a levels

pypy3 main.py --{GAME} --{ALGORITHM} --generate-corpus

Plot the MAP-Elites grid

python3 main.py --{GAME} --{ALGORITHM} --plot-map-elites

Run level generation multiple times

pypy3 main.py --{GAME} --{ALGORITHM} --average-generated --runs {n}

Plot average generated

python3 main.py --{GAME} --{ALGORITHM} --plot-counts

Special Case

Since the paper, I've gone back and made some changes to clean up the code. When I wrote this however long ago, I had this idea that a classless config as module was a good idea... it wasn't. But, we live and we learn. So, I've added some interfaces to make the code easier to follow.

I've also added Match3. The problem with Match3 is that I implemented it for the web, so it's all in TypeScript. To make it callable by Python, I wrote a server that can be run locally. So, if you want to run Match3 yourself, you first have to get the server running, which is pretty simple and the instructions are in the Match3 GitHub, sorry for any inconvenience.

Citation

@inproceedings{biemer2021gram,
  title={Gram-Elites: N-Gram Based Quality-Diversity Search},
 author={Colan Biemer and Alejandro Hervella and Seth Cooper},
 booktitle={Proceedings of the FDG workshop on Procedural Content Generation},
 year={2021}
}