How to Run the Program:

Python (http://www.python.org) and pygame 
(http://www.pygame.org) will have to be installed on
the computer before this program can run. Unzip the 
contents somewhere onto your computer. Find the file 
labeled "main.py" and run it. 

Controls:

 - Close the window to exit.
 - Press escape to exit.

Goal:

The goal is for the computer to match the image 
through a genetic algorithm. The user need only
sit back and watch.

This program can run up to near 7 minutes and
usually find a match when the number of generations
reaches the 900s, though it is not guaranteed.

Features:

 - This program takes a given image, face.png, and 
generates 1000 generations of 500 entities of 
images made up of random colors. Through mating
and mutating, the generations evolve and work
closer towards looking like the given image.

 - The best image of every generation is stored
as a .bmp file in the results folder. If images are
there from a previous run, they will be overwritten.

 - The algorithm uses a tournament selection, which 
mean 10 random entities from every generation are
chosen to mate until 40% of a new generation is 
made. The remaining 60% is chosen at random to
continue onto the next generation from the
current one.

 - There is another method programmed where the 
entities are sorted and the ones chosen to mate are
picked at random, but inclined to be picked from the
front (the best). However, this method is not used
as it does not yield as good results.

 - An interesting but unintended feature is that
although a pixel that matches the given picture
is allowed to mutate or change through mating, 
it does not when using the tournament method. This
can only be explained by the way the early samples
are chosen in the first generations.

Resources:

 - Pygame was used to render the screen and save 
screenshots of the progress made. 
It is licensed under LGPL 
(http://www.gnu.org/copyleft/lgpl.html).