Evolution simulator

Created by Cary Huang and released on OpenProcessing.org. Many thanks for creating this educational and entertaining gem of code!

Some small functionality added by Aron van Ammers.

Up to now I have been unable to get in contact with the original author Cary Huang. Cary, if you read this, I'd be very happy to move this code over to a repo owned by you with proper attribution of the code. Just message me here.

License

Unknown!

Introduction

Original video explanations by Cary Huang

Prerequisites

Install Processing v1.5.1.

Note that the code currently is NOT compatible with the more recent Processing versions 2 and 3. You'll get an error message "GraphImage cannot be resolved or is not a field" when running. Happy to accept a pull request for anyone making it compatible with Processing 3.

Running

Open the .pde file in Processing and run it.

Configuration: changing the simulation

The top of the file contains some easy to edit variables that influence the simulation. They are documented in the file.

Some interesting ones:

Randomization and mutations

• USE_RANDOM_SEED - when true, the simulation is deterministic, i.e. it will give exactly the same results if you would run it again.
• SEED - in combination with USE_RANDOM_SEED, this determines the run. The same number will lead to the same results.
• MUTABILITY_FACTOR - how fast creatures mutate. Lower values will lead to a monoculture of highly optimized creatures. Higher values give some more variety, but also to a lot of "mishaps" because of nonsuccessful mutations.

Minimum complexity

• CREATURE_MIN_NODES - the minimum amount of nodes that creatures must have. Allows for requiring more complex creatures.
• CREATURE_MIN_MUSCLES - the minimum amount of muscles that creatures must have.

Randomization of obstacles

• RANDOMIZE_RECTANGLES - if true, the rectangles (obstacles) will be randomized over each generation. This will give more versatile creatures an advantage, i.e. creatures that aren't completely optimized to climb a single stairway with identical stairs.
• RECTANGLES_PROGRESSIVE_MUTATION - if true, randomizes the rectangles slightly over each generation. Think of this as a natural environment that slightly changes over time. If false, the rectacles will be completely randomized (from their original configuration) for each generation.
• RECTANGLES_MUTABILITY_FACTOR - sets how much the rectangles are randomized.