johnlime / pocs_paper

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Zipf’s Law and CCDF Analysis on Deep Reinforcement Learning

Supplementary code for final paper in CS6701-A: Principles of Complex Systems at University of Vermont

Paper

Get Started

Python and Anaconda are required. MacOS or Linux environment is recommended.

Execute the following to install all dependencies in a virtual environment.

conda env create -f environment.yml
conda activate rl_power_law

Add the current directory to Python Path.

export PYTHONPATH=$PWD

Generate dataset for OpenAI Gym "CartPole-v1" (This will overwright the preexisting dataset included in the repository at results_cartpole-v1).

python run.py -h
python run.py

Open Power Law Analysis.ipynb in Jupyter Notebook for Zipf distribution and CCDF analysis.

About


Languages

Language:Jupyter Notebook 90.2%Language:Python 9.8%