Reasoning-Agents-Project

Project for the course of Reasoning Agents at University La Sapienza of Rome A.Y. 2020/2021.

The goal of the project is to implement solvers for different types of games on graphs namely:

• Reachability
• Safety
• Büchi
• Co-Büchi
• Parity

Description

The user can solve a game by passing as input a game arena as a .txt file with the format shown in the figure:

In the main folder is possible to find the both the report the slides that explain in detail the implementation of the algorithms.

How to run the code

Note : players are called player 0 and player 1 You can try the tool by passing as argument at least '--game' since there are some default values for the arena and the target set.

Here we list some example of commands to run the code:

• To show the help diag: solver.py -h
• To solve a game: solver.py --game` passing the corresponding game:
  • --game reachability for reachability games
  • --game safety for safety games
  • --game Bübuchi for Büchi games
  • --game cobuchi for Co-Büchi games
  • --game parity for Parity games
• To specify the arena to solve add the flag: --arena assets/arena0.txt with the path that brings to the .txt file containing the arena to solve
• To specify the target set add the flag: --target 4, 6 with the desired target set (here nodes 4 and 6)