Chess player simulator

Train and simulate the chess player's style

This project is designed to simulate the playing style of a specific chess player. It achieves this by first downloading historical matches of the given player from Chess.com. These matches are then used to train a machine learning algorithm. Once trained, the algorithm can be used to predict and simulate the player's behavior in various chess scenarios.

Note: Notice that the project contains a requirements.txt file.

Downloader

The downloader.py script is used to download chess games of a particular player from Chess.com and store them as PGN (Portable Game Notation) files. The script takes the following command-line arguments:

• --username: The username of the chess player whose games you want to download. Defaults to "Joseda8".
• --year: The year of the games to download. Defaults to the current year.
• --remove: If specified, removes existing PGN files in the matches folder before downloading new ones.
• --folder: The folder where the PGN files will be stored. Defaults to "matches".

To use the script, navigate to the project folder in a terminal and run the following command:

python downloader.py --username <player_name> --year <year> --remove --folder <folder_name>

For example, to download games from the year 2023 for the player "Joseda8" and store them in the "matches" folder, run the following command:

python downloader.py --username Joseda8 --year 2023

Chess player

ChessStudent

The ChessStudent class is used to train a machine learning model that predicts chess moves based on historical games. The class supports two different machine learning algorithms for predicting moves: k-Nearest Neighbors (k-NN) and Random Forest. The trained model can be used to simulate the playing style of a specific chess player.

ChessMatch

The ChessMatch class is used to play a game of chess against the bot trained by the ChessStudent class. It generates moves for the bot based on the trained model and plays the game until a game-over state is reached.

The script takes the following command-line arguments:

• --games_directory: The directory where the PGN games files are stored.
• --player_name: The name of the player that the bot should learn moves from.
• --cache: Indicates whether the results should be cached or read from the cache.
• --bot_color: The color that the bot is playing as (either "white" or "black"). Defaults to white.
• --algorithm: The machine learning algorithm to use (either "knn" or "rf"). Defaults to knn.

To use the script, navigate to the project folder in a terminal and run the following command:

python chess_match.py --games_directory <path_to_games_directory> --player_name <player_name> --cache --bot_color <bot_color> --algorithm <algorithm>

For example, to play a game of chess against the bot trained on games from the "matches" folder for the player "Joseda8", with the bot playing as white and using the k-NN algorithm, run the following command:

python chess_match.py --games_directory matches --player_name Joseda8 --cache --bot_color white --algorithm knn