Merge branch 'dev'

README.md

@@ -1,21 +1,74 @@
 # Tetris AI
 
+<div align="center">
+
+![GitHub Workflow Status (with event)](https://img.shields.io/github/actions/workflow/status/CogitoNTNU/TetrisAI/ci.yml)
+![GitHub top language](https://img.shields.io/github/languages/top/CogitoNTNU/TetrisAI)
+![GitHub language count](https://img.shields.io/github/languages/count/CogitoNTNU/TetrisAI)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+[![Project Version](https://img.shields.io/badge/version-1.0.0-blue)](https://img.shields.io/badge/version-1.0.0-blue)
+
+![logo](docs/img/Logo.webp)
+
+</div>
+
 ## Description
 
-<!-- TODO -->
+This project is our attempt at making an AI that can play Tetris. First of all we made the environment (the game itself) and then we started working on multiple AI's that can play the game. The AI's are based on different algorithms and strategies. We have implmented:
+
+- Random agent
+- Heuristic agent with set weights
+- Genetic algorithm to find the best weights for the heuristic agent
+
+The game is playable/viable both in the terminal and in a GUI. The GUI is made with Pygame.
+
+## How to run and install
+
+- Have python installed (tested with 3.10+ but should work on older versions as well)
+- Have pip installed
+- Clone the repository
+- Set up a virtual environment (optional but recommended) see [here](docs/guide/venv.md) for a guide
+- Install the required packages with pip
+
+```bash
+pip install -r requirements.txt
+```
+
+- Done! You are now ready to run the game
+
+## Usage
+
+To play the game yourself, run the following command:
+
+```bash
+python main.py play
+```
+
+To let the agent play the game, run the following command:
+
+```bash
+python main.py agent <agent>
+```
+
+where `<agent>` is the agent you want to use. The available agents are: `random`, `heuristic`, `genetic`
 
-## Setup
-### Prerequisites
-- Ensure that git is installed on your machine. [Download Git](https://git-scm.com/downloads)
-- Ensure Python 3.12 or newer is installed on your machine. [Download Python](https://www.python.org/downloads/)
+To train the genetic agent, run the following command:
 
-### Clone the repository
 ```bash
-git clone https://github.com/CogitoNTNU/TetrisAI.git
-cd TetrisAI
+python main.py train
 ```
 
-## Contributors
+## Testing
+
+To run the test suite, run the following command from the root directory of the project:
+
+```bash
+python -m pytest
+```
+
+## Team
+
+The team behind this project is a group of students at NTNU in Trondheim, Norway, developed during the spring semester of 2024. The team consists of:
 
 <table align="center">
     <tr>
@@ -63,3 +116,5 @@ cd TetrisAI
         </td>
     </tr>
 </table>
+
+![Group picture](docs/img/Team.png)

demo.py

@@ -0,0 +1,9 @@
+from src.agents.agent_factory import create_agent
+from src.game.tetris import Tetris
+from src.game.TetrisGameManager import *
+
+if __name__ == "__main__":
+    board = Tetris()
+    agent = create_agent("heuristic")
+    manager = TetrisGameManager(board)
+    manager.startDemo(agent)

main.py

@@ -1,54 +1,52 @@
+import argparse
 from src.game.tetris import Tetris
-from src.game.TetrisGameManager import *
-from src.agents.agent import Agent, play_game
+from src.game.TetrisGameManager import TetrisGameManager
 from src.agents.agent_factory import create_agent
-from src.agents.heuristic import (
-    utility
-)
-from src.agents.heuristic_trainer import train
 from src.agents.geneticAlgAgentJon import GeneticAlgAgentJM
-
-def test():
-    # algAgent = GeneticAlgAgentJM()
-    # algAgent.number_of_selection(2)
-    # print(algAgent.getBestPop())
-
+def self_play():
+    """Start a self-playing Tetris game."""
     board = Tetris()
-    agent = create_agent("heuristic")
+    manager = TetrisGameManager(board)
+    manager.startGame()
+def demonstrate(agent_type: str):
+    """Demonstrate gameplay with a specified agent."""
+    board = Tetris()
+    agent = create_agent(agent_type)
     manager = TetrisGameManager(board)
     manager.startDemo(agent)
-
+def train_genetic_algorithm(population_size: int = 10):
+    """Train the genetic algorithm agent."""
+    alg_agent = GeneticAlgAgentJM()
+    alg_agent.number_of_selection(population_size)
+    print(alg_agent.getBestPop())
+def main():
+    """Main entry point to handle command-line arguments."""
+    parser = argparse.ArgumentParser(description="Tetris Game with different options.")
+    subparsers = parser.add_subparsers(dest="command", help="Sub-command help")
+    # Self-play parser
+    subparsers.add_parser("play", help="Start a self-playing Tetris game.")
+    # Demonstrate parser
+    demonstrate_parser = subparsers.add_parser(
+        "agent", help="Demonstrate gameplay with a specific agent."
+    )
+    demonstrate_parser.add_argument(
+        "agent", type=str, help="Agent type for demonstration."
+    )
+    # Genetic algorithm training parser
+    train_parser = subparsers.add_parser("train", help="Train the genetic algorithm agent.")
+    train_parser.add_argument(
+        "--population_size", type=int, default=10, help="Population size for the genetic algorithm."
+    )
+    # Parse the arguments
+    args = parser.parse_args()
+    # Route commands to the appropriate functions
+    if args.command == "play":
+        self_play()
+    elif args.command == "agent":
+        demonstrate(args.agent)
+    elif args.command == "train":
+        train_genetic_algorithm(args.population_size)
+    else:
+        parser.print_help()
 if __name__ == "__main__":
-
-    # game = Tetris()
-    # agent: Agent = create_agent("heuristic")
-    # sum_rows_removed = 0
-    # for i in range(10):
-    #     end_board = play_game(agent, game, 7)
-    #     end_board.printBoard()
-    #     sum_rows_removed += end_board.rowsRemoved
-
-    # print(f"Average rows removed: {sum_rows_removed / 10}")
-
-    # possible_moves = game.getPossibleBoards()
-    # for boards in possible_moves:
-    #     print(utility(boards, 0, -1, 0, 0, 0))
-    #     boards.printBoard()
-
-    # board = Tetris()
-    # manager = TetrisGameManager(board)
-    # agent = create_agent("heuristic")
-
-    # # manager.startGame()
-
-    # # train()
-
-
-    # algAgent = GeneticAlgAgentJM()
-    # algAgent.number_of_selection(2)
-    # print(algAgent.getBestPop())
-
-    test()
-
-
-    # cProfile.run('main()', 'restats')  
+    main()

requirements.txt

@@ -1,8 +1,6 @@
-colorama==0.4.6
 iniconfig==2.0.0
+numpy==1.26.4
 packaging==24.0
-pluggy==1.4.0
+pluggy==1.5.0
 pygame==2.5.2
-pynput==1.7.6
-pytest==8.1.1
-six==1.16.0
+pytest==8.2.0

src/agents/geneticAlgAgentJon.py

@@ -79,7 +79,9 @@ def play_game(self, agg_height, max_height, lines_cleared, bumpiness, holes):
 
                 move += 1
             # Advance the game by one frame
-            board.doAction(Action.SOFT_DROP)
+                board.doAction(Action.SOFT_DROP)
+                if board.blockHasLanded:
+                    board.updateBoard()
             #board.printBoard()
 
             total_cleared += board.rowsRemoved
@@ -103,7 +105,8 @@ def replace_30_percent(self, pop_list: list[list[list[float], float]]) -> list[l
      # TODO create method for fetching a random 10%, and finds the two with highest lines cleared, and makes a child (with 5% chance of mutation)
     def paring_pop(self, pop_list: list[list[list[float], float]]) -> list[list[float], float]:
         # Gets the number of pops to select
-        num_pops_to_select = int(len(pop_list) * 0.1)
+        # num_pops_to_select = int(len(pop_list) * 0.1)
+        num_pops_to_select = int(len(pop_list) * 0.5)
 
         # Get a sample of pops based on the previous number
         random_pop_sample = random.sample(pop_list, num_pops_to_select)
@@ -113,13 +116,18 @@ def paring_pop(self, pop_list: list[list[list[float], float]]) -> list[list[floa
 
         # Gets the child pop of the two pops
         new_pop = self.fitness_crossover(highest_values[0], highest_values[1])
-
+        norm = np.linalg.norm(new_pop[0])
+        if norm == 0:
+            norm = 1e-10  # or some small constant
+        new_pop[0] = [i / norm for i in new_pop[0]]
+        new_pop[0] = [i / norm for i in new_pop[0]]
+
         # Mutate 5% of children pops
-        if random.randrange(0,1000)/1000 < 0.2:
+        if random.randrange(0,1000)/1000 < 0.05:
             random_parameter = int(random.randint(0,4))
             new_pop[0][random_parameter] = (random.randrange(-200, 200)/1000) * new_pop[0][random_parameter]
 
-        new_pop[0] = (new_pop[0] / np.linalg.norm(new_pop[0])).tolist()
+        #new_pop[0] = (new_pop[0] / np.linalg.norm(new_pop[0])).tolist()
 
         return new_pop
 

src/agents/random_agent.py

@@ -7,8 +7,8 @@
 class RandomAgent(Agent):
     """Random agent that selects a random move from the list of possible moves"""
 
-    def result(self, board: Tetris) -> Action:
+    def result(self, board: Tetris) -> list[Action]:
         possible_actions = get_all_actions()
-        return choice(possible_actions)
+        return [choice(possible_actions)]
 
 

src/game/tetris.py

@@ -7,7 +7,7 @@
 
 from src.game.block import Block
 
-DEMO_SLEEP = 0.05
+DEMO_SLEEP = 0
 
 
 class Action(Enum):