Artificial life project based on physical simulation and neural networks in python.
Agents (blue) called Goopies are placed in an environment where they have to eat food (green) to survive. If they eat enough, they are able to reproduce, with a slight chance of mutations occurring. This combination of mutation and selection creates the conditions for the evolution of traits, which leads to agents more suited to the environment.
Full simulation:
Close-up of a single evolved Goopie (the white circle is the vision radius):
To install all the required dependencies, simply clone the repository and install using pip in your virtual environment.
N.B.: Sometimes pyglet can cause problems if installed directly in system packages, so using a virtual environment is highly recommended.
python -m venv .venv
source ./venv/bin/activate
Then install all the requirements:
pip install -r requirements.txt
To run the project simply execute main.py:
python src/main.py