Skip to content

Useful code for running Reinforcement Learning experiments.

License

Notifications You must be signed in to change notification settings

AmiiThinks/rltoolkit

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RLtoolkit

RLtoolkit Version 1.0 b7 October 27, 2017

Note: the arrangement of modules in this package is subject to change (and most likely will!) with future releases.

##To use:

  1. Move the RLtoolkit folder to your site-packages folder for Python OR export PYTHONPATH=$PYTHONPATH:/path/to/rltoolkit

  2. You can then import as you would any other package

    In Python: from RLtoolkit import Tiles.tiles as tiles from RLtoolkit.traces import * from RLtoolkit import * etc. From the command line: python -m RLtoolkit.gridworld.gwDemoG

##Contents:

examples - a folder of demos to run, using the toolkit

  • the maintenance example
  • mountain car (GUI and nonGUI)

fa - function approximation demo (currently only uses tile coding) gridworld - gridworld demo G - a general low level graphics drawing package Quickgraph - a simple graphing package (uses g) for 2 and 3d graphs Tiles - the tile coding package

basicclasses.py - definitions for simulation, agent and environment objects RLinterface - contains the rl interface object and methods for it traces - eligibility traces handling utilities - some general utilities guiwindow - a generic simulation window with buttons and menus (not yet documented for users) demo - helps load the demos and run them guiuser - loads all the major tools from the toolkit (if you do from guiuser import *, you get all the tools with no need to prefix module names)

These are shortcuts to some of the tool packages: tiles - imports Tiles.tiles g.py - imports G.g graph.py - imports Quickgraph.graph graph3d.py - imports Quickgraph.graph3d

There are some extra things for the RLinterface: RLinterface2 is the interface described on the first RLAI web page as RL 6 RLinterface3 is a start of the interface described on the RL benchmarks page rlitest2a is a simple test of RLinterface2 rlitest2b is the random walk example with RLinterface2 rlittest3a is a simple test of RLinterface3 rlitest3b is the random walk example with RLinterface3

About

Useful code for running Reinforcement Learning experiments.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published