Setup can be done with globally installed packages (default way to setup a python project) or can be setup in a virtual environment (cleaner / easier to maintain, but more up-front work).
To setup in the standard way:
pip3 install tensorflow tensorflow_probability numpy matplotlib
To setup in a virtual environment:
pip3 install virtualenv
cd Path/to/project/directory
virtualenv -p python3 659env
To "login" to the virtualenv before running the code:
. 659env/bin/activate
Finally, install the packages:
pip install tensorflow tensorflow_probability numpy matplotlib
To run the code, first activate the virtual environment (if you are using one). Then run:
python q_learning.py