REINFORCEMENT LEARNING ASSIGNMENT 2

Hi all, I've added the two notebook files to the directory 'src'. We can put all our code files in there.

We could make another folder for the report/presentation, but also keep them in this repository so we can all make changes and collaborate.

Getting started

If not already installed, download git from https://git-scm.com/downloads and go through the installer.

If not done so already, you will need to set up an ssh key, which github will use to authenticate you. See here: https://docs.github.com/en/authentication/connecting-to-github-with-ssh

Open up a terminal (git bash on windows), navigate to a folder where you want to store the files and run the command "git clone git@github.com:Le5tes/robot-reinforcement-learning.git"

This should create a directory called robot-reinforcment-learning, with this readme and our code files inside.

Links

Gym jiminy code: https://github.com/duburcqa/jiminy

Gym jiminy documentation: https://duburcqa.github.io/jiminy/

• not great documentation

OpenAI Gym documentation: https://www.gymlibrary.ml/

• gym jiminy implements this API - the methods documented here are the same methods as GJ uses - so this is very useful

Keras documentation: https://keras.io/api/

• we will need to implement some neural nets so lets use keras?

Useful commands

Shell commands

Here are some useful commands for navigating when using a bash terminal (e.g. git bash)

• pwd
  • Shows which directory you are in
• cd
  • moves to the directory
• ls
  • Lists the files in the directory you are currently in

Git commands

These are the basic git commands that will be most useful. There are many more!

• git pull

  • Retrieves the latest code from github

• git add .

  • stages your changes for commiting

• git commit -m "some message"

  • makes a commit - a record of the state of the code with a message

• git push

  • pushes your commits to github