Artificial Intelligence (For Dummies): Q-Learning

Preparation Instructions

This coding dojo uses Python, version 2.7 to be precise. If you're experienced using Python and you have a working Python 2.7 setup configured, please feel free to use that. Otherwise, you can follow theses instructions to prepare for the coding dojo, so that we can get started right away.

• Install Miniconda. On macOS you can also use homebrew if you have that installed: brew cask install miniconda. Verify that it works and is up to date. You can do that in a terminal by running:

  $ conda update -y conda
  [... lots of output ...]
  Preparing transaction: done
  Verifying transaction: done
  Executing transaction: done
  $ conda --version
  conda 4.8.2
  $ conda init <your shell, eg zsh or bash>
  [...]
  modified      /Users/nlv19703/.zshrc
  [...]

• Close the terminal and open a new one for the changes to take effect.

• Clone the repository that we will be using for the coding dojo: git clone https://github.com/Mocolate/reinforcement_learning_pacman.git.

These setup instructions assume that you'll be using PyCharm Community Edition. (If you want to use something else, make sure you know how to get everything up and running using that IDE/Editor.) After installing Pycharm, follow these steps:

• In the Welcome screen, choose 'Create New Project'.

• Pick a folder where you want to store the project, and a new folder named qlearning inside that folder.

• Expand the "Project Interpreter: New Conda environment" settings.

• Make sure "New environment using: Conda" is selected.

• Change the Pyton version to 2.7.

• Click "Create".

• PyCharm will create the Conda environment; this may take a while.

• In the steps above, you cloned a repo. It has a folder reinforcement. Copy all files and folders inside of that to the new project folder you just created. So there should now (amongst others) be a file requirements.txt inside your qlearning folder.

• Open a terminal inside PyCharm. This should cause the conda environment to be automatically activated inside that terminal. If you're not sure: conda info should say that the active environment is qlearning. (If it's not active, run conda activate qlearning.)

• Run the following command to install the dependencies:

  $ pip install -r requirements.txt
  [...]
  Successfully installed atomicwrites-1.3.0 attrs-19.3.0 configparser-4.0.2 contextlib2-0.6.0.post1 funcsigs-1.0.2 importlib-metadata-1.5.0 more-itertools-5.0.0 packaging-20.1 pathlib2-2.3.5 pluggy-0.13.1 py-1.8.1 pyparsing-2.4.6 pytest-4.6.9 scandir-1.10.0 six-1.14.0 wcwidth-0.1.8 zipp-1.1.0

• Verify if your setup if fully functional by running:

  $ pytest tests/test_if_it_works.py
  [...]
  ========================= 1 passed in 0.03s =========================

  If you get errors at this point, you should verify that you are using the correct version of python and pytest. In the output of pytest you can see both. If this is incorrect, run which pytest to see which one you are using. It should be the one inside the conda qlearning env; if not, run rehash or open a new terminal and try again. If that still fails, make sure that conda comes first in your path.