sunshineluyao / eth-rl

eth-el

Replicable code for "Redesign Incentives in Proof-of-Stake Ethereum: An Interdisciplinary Approach of Reinforcement Learning and Mechanism Design"

How to run the code

• Clone the repository to your local directory
• Install or upgrade pip: https://pip.pypa.io/en/stable/installation/

pip install --upgrade pip

• Run the code to install required packages

pip install -r requirements.txt

Explanation of the files

• requirements.txt including all the installed packages is generated using the following code:

pip freeze > requirements.txt

• the "gym" folder includes the Reinforcement Learning (RL) environment
• the "model" folder includes the RL algorithms
• the "tmp" folder stores the best RL algorithm
• the "results" folder stores the output data from trained RL
• the "analysis.ipynb" represents the virtualizations
• the "figures" folder saves the output figures from "analysis.ipynb"
• the "makefile" defines the short-cut of pytest command