OBSOLETED

This package is obsoleted by the newer and updated package at JuliaReinforcementLearning/ArcadeLearningEnvironment.jl. Please use the updated package instead of this one.

ArcadeLearningEnvironment.jl

This package is a Julia wrapper for the ArcadeLearningEnvironment (ALE).

ALE is a modified emulator for the Atari 2600 that can emulate around 50 games with additional access to game state information and in-game rewards. This is useful for learning and benchmarking artificial intelligence agents playing computer games.

A higher-level interface on top of this package is available thanks to Tom Breloff, at the AtariAlgos.jl repository.

Citation

If you use this package for research publications, please cite the following paper to acknowledge the work that went into ALE.

@ARTICLE{bellemare13arcade,
	author = {{Bellemare}, M.~G. and {Naddaf}, Y. and {Veness}, J. and {Bowling}, M.},
	title = {The Arcade Learning Environment: An Evaluation Platform for General Agents},
	journal = {Journal of Artificial Intelligence Research},
	year = 2013,
	month = 06,
	volume = 47,
	pages = {253--279}
}

Installation

On Mac OS and Linux the package automatically downloads and builds version 0.5.1 of the ArcadeLearningEnvironment by issuing the following commands.

Pkg.clone("https://github.com/nowozin/ArcadeLearningEnvironment.jl.git")
Pkg.build("ArcadeLearningEnvironment")

On Windows (which I have not tried yet) you can build the libale_c.dll file manually and set the LIBALE_HOME environment variable to the directory containing this file. Then, the above two commands should work as well. Note that this is untested and any correction or feedback is welcome.

Example

using ArcadeLearningEnvironment

# For this example you need to obtain the Seaquest ROM file from
# https://atariage.com/system_items.html?SystemID=2600&ItemTypeID=ROM

episodes = 50

ale = ALE_new()
loadROM(ale, "SEAQUEST.BIN")

S = Array(Int, episodes)
TR = Array(Float64, episodes)
for ei = 1:episodes
    ctr = 0.0

    fc = 0
    while game_over(ale) == false
        actions = getLegalActionSet(ale)
        ctr += act(ale, actions[rand(1:length(actions))])
        fc += 1
    end
    reset_game(ale)
    println("Game $ei ended after $fc frames with total reward $(ctr).")

    S[ei] = fc
    TR[ei] = ctr
end
ALE_del(ale)