Environments, similar to labeled datasets in supervised learning, are the essential part of RL as they dictate the information that has to be learned and the choice of algorithms. In this section, we'll take a look at the main differences between the types of environments and list some of the most important open source environments.