DOCUMENTATION.txt

V 1.1.5

Executable files:
	main_da.py
		Analyse a dataset to collect statistics about the state composition (
	
	main_dl.py
		Analyse a dataset to collect statistics about the legal moves for each states
		
	main_play.py
		Client that can be used to play the game of Nine Men's Morris.
		Communicate thought socket on port 5802 for white player and 5803 for black player.
		A state string is expected as input on the port and a move string is sent back with format "TO[FROM][REMOVE]".
	
	main_reliability.py
		Execute a reliability test: it measure the ability of the system to assign an higher score to legal choices.
	
	main_test.py
		Execute a legality and accuracy test: it measures the ability of the system to provide accurate moves (equal to the ones in the dataset) abd legal moves.
	
	main_training.py
		Allows to build a network from scratch or to load an existing one and to train it.
		Networks characteristics must be defined modifying the file.


Source files:

	dataprocessing.py
		Contains the functions used to manipulate the data.
		
	legality.py
		Contains the functions used to establish the legality of a move.
		
	networks.py
		Contains the functions used to create, load, save and use neural networks.
		
	testing.py
		Contains the functions used to test the system according to different characteristics.
		
	training.py
		Contains the functions used to train a neural network system.

		
Changelog:
v1.1.5
	The loading of dataset for training has been improved