GitHub - lorenzobini/comparative-system-multiagent-algorithms: Comparative system for multiagent algorithms with different learning strategies. The analysis is carries with the helps of a nash equilibria comparison, the replicator dynamic and a simple grand table with the average reward obtained.

lorenzobini / comparative-system-multiagent-algorithms Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Comparative system for multiagent algorithms with different learning strategies. The analysis is carries with the helps of a nash equilibria comparison, the replicator dynamic and a simple grand table with the average reward obtained.

0 stars 0 forks Branches Tags Activity

Star

Notifications

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 33 Commits
.idea		.idea
Game.py		Game.py
GrandTable.py		GrandTable.py
Main.py		Main.py
MatrixSuite.py		MatrixSuite.py
Nash.py		Nash.py
README.txt		README.txt
ReplicatorDynamic.py		ReplicatorDynamic.py
Strategies.py		Strategies.py
Utils.py		Utils.py
gambit-enummixed.exe		gambit-enummixed.exe

Repository files navigation

The program makes use of the Python libraries numpy, matplotlib.
The program needs Gambit to compute Nash equilibria correctly. If not installed, download and install
Gambit 15:
http://www.gambit-project.org/

Copy the gambit-enummixed.exe executable from the installation folder to the folder containing the Nash.py script.

To store the images from the replicator dynamic, create a folder "/img" inside the folder containing the
Main.py script.
The method game_session() in Main.py contains a call to to_latex() of GrandTable.py. This method converts the
resulting grand table of the game into a formatted table suitable for LaTeX. To store the tables, create a
folder "/grand_table" inside the folder containing the Main.py script.

Run Main.py to play the games and compute the statistics.