readme.md
In hosscine/rlenv: Provide environment for RL and GUI of the environment via shiny

rlenv

This is r package of the environment for the reinforcement learning to explore a volcano. You can explore the environment with shiny application that is included in the package.

demo

devtools::install_github("hosscine/rlenv")

To start GUI for manualy exploring:

library(rlenv)
startEnvironment()

To designe a reinforcement learning agent for the environment:

library(rlenv)
env <- volcanoExplorer$new()

A instance of the volcano explorer has methods below: initialize() * resets position and velocity of the agent observeContinuous() * observes state(position and velocity) continuously observeDiscrete() * observes state(only position) discreately actionContinuous(action.x, action.y, torque = 2) * affects continuously 2D action to environment actionDiscrete(discrete.action, torque = 2) * affects discretely 1D action to environment * discrete.action is required (0, 1, 2, 3, 4) to (stop, move up, move down, move left, move right), respectively getReward() * gets reward of current state * plot() * plots the environment and the position of the agent