WOLFRAM

Wolfram Language & System Documentation Center
Wolfram Language Home Page »

"OpenAIGym" (Reinforcement Learning Environments)

[Experimental]

"OpenAIGym" provides an interface to the Python OpenAI Gym reinforcement learning environments package.

To use "OpenAIGym", the OpenAI Gym Python package must be installed.

The OpenAI Gym Python package is only officially supported on Linux and macOS platforms. Additionally, several different families of environments are available. Examples on this page use the "Atari" family of environments. Depending on the host system details, this family can be installed from the command line using one of the following (additional detailed installation instructions are available here):

$ pip install "gym[atari]"
$ pip3 install "gym[atari]"
$ pip install gym atari-py

ExternalEvaluate is used to interface with this Python package. It is up to the user to ensure that ExternalEvaluate is set up with a Python environment with Gym installed. A simple test to ensure that Gym is correctly set up is to run the following:

ExternalEvaluate["Python", "import gym"]
CarRacing-v0 environment from the OpenAI Gym

Examples

Basic Examples  (2)Summary of the most common use cases

Open the "Atlantis-v0" Atari environment:

In[1]:=1
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-4d4cil
Out[1]=1

Observe the initial state:

In[2]:=2
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-iudoxc
Out[2]=2

In this case, the "ObservedState" is an array of pixel values. Visualize this as an image:

In[3]:=3
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-rgq8j8
Out[3]=3

Obtain a random action:

In[4]:=4
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-h6rdm8
Out[4]=4

Take the random action:

In[5]:=5
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-mikjjd
Out[5]=5

Taking an action usually modifies the "ObservedState":

In[6]:=6
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-mvwn53
Out[6]=6

Reset the environment to an initial state. The initial "ObservedState" is returned:

In[7]:=7
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-9jaj26
Out[7]=7

Close the environment:

In[8]:=8
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-pjakth

Open the "Breakout-v0" environment:

In[1]:=1
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-mygsbe
Out[1]=1

Visualize a random agent playing Breakout:

In[2]:=2
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-bci0ly
Out[2]=2

Close the environment:

In[3]:=3
✖
https://wolfram.com/xid/0d2t83q23q3v678u5nh3gb-8e6xbw