Visual Studio Code will be used as the IDE for this project, though most any text editor will work for reproducing it. The project has the following dependencies:

• Git
• Python 3 environment with the following packages installed:
  • PyGame; open source game development platform for Python
  • Scipy; open source scientific computing package suite for Python
  • Scikit-Learn; popular machine learning library for Python
  • Keras; expansive open source library for neural networks
  • TensorFlow; suite of open source deep learning libraries
  • OpenCV; open source image processing library

Installing a Python distribution and creating the environment will be done through Anaconda. To automatically create a flappy environment preconfigured with the correct dependencies for this project:

1. Download the requirements file environment.yml to a {path} on your system

2. Edit environment.yml so prefix matches your Anaconda \envs path

3. From Anaconda Prompt run:

   conda env create -f {path}\environment.yml

There are several widely available opensource codebases in existence that emulate Flappy Bird and provide all necessary assets. This project will utilize a popular clone FlapPy Bird, built in Python using pygame and available on GitHub under the MIT License , which provides all functionality and should require relatively little code modification to make it suitable for an RL agent.

1. From a command terminal cd to the project directory and run:

   git clone https://github.com/sourabhv/FlapPyBird.git

2. All graphics are located in the \assets folder

3. The game can be launched by running flappy.py from your Python terminal

4. The ↑ or Space keys are used to flap in lieu of tapping on the screen