Skip to content

l3robot/superchess

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

28 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

superchess - a super simple chess engine

Are you a user or a developer?

For users

There's only two simple steps in order to play against our chess engine.

1. Installing

Installation is as simple as

pip install -r requirements.txt .

2. Playing

To play a game against an alphabeta of depth3 opponent, simply type

python3 examples/qtgui.py

The pieces need to be dragged (click mouse on the piece's current square, and lift mouse button on goal square).

If for any reason the gui experiences problems, once can also use a simpler keyboard-based input method jupyter notebooks (see the gui section for more info). First download jupyter, then open the examples/play_against_computer.ipynb notebook:

cd examples
jupyter notebook
# open play_against_computer.ipynb through the jupyter API

and play!

play-vs-computer

Moves have to be entered manually, using either of

  • Standard Algebraic Notation (e4, Nf3, exd5, etc.)
  • Universal Chess Interface (e2e4, g1f3, e4d5, etc.)

For developers

Installation

As a developer it's better to install superchess using

pip install -e -r requirements.txt .

This will keep track of the changes you make to files, and you won't need to reinstall superchess after every change.

Unit Testing

Unit tests can be run from the tests folder by running

python3 -m unittest

GUI

We use python-chess as a frontend, which generates svg images, and so the GUI needs to show this svg image and then ask the user for move input. There are two GUIs implemented, one for which the input is click-based and another for which it is keyboard-based.

The mouse/click-based GUI is the main one, and is implemented using PySide2. The code can be found in examples/qtgui.py.

The keyboard-based gui uses IPython's Display function, and needs to be ran inside a jupyter notebook.

alpha-beta pruning misses checkmates in 1!

Can you find the checkmate in 1?

Checkmate in one rooks

The classical alphabeta algorithm with a depth of 3 misses it! This is because moving the King or Rook on h1 first still leads to a checkmate, and alphabeta doesn't distinguish between checkmates that happen in one move or those that happen in three! This is why our implementation alphabeta_treesearch returns (best_score, depth, best_moves), where depth is used to distinguish between moves with the same score that otherwise look equivalent to alphabeta.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages