Typing Monkeys and Genetic Algorithms

Recently I spent time playing with generic algorithms. I was looking for something easy enough to implement (read: PLAIN LAZY) and I found the Hello World! (literally) of genetic algorithms.

Gathering the algorithm itself from the C++ example (in the link above), I re-designed and re-coded the whole thing in C#.

TypingMonkey is a very simple application which uses a number (defined by the user) of random strings to evolve the user input text.

From the screen-shot below you can see how it took the algorithm 87 generations to evolve the string ".NET Butchering" from an initial population of 2500 random 15-characters strings:

Why "typing monkey"? Inspired by the Infinite Monkey Theorem, I am pretending our population of random string is generated by a typing monkey:

/// The Typing monkey generates random strings - can't be static 'cause it's a monkey.
/// If you wait long enough it will eventually produce Shakespeare.
class TypingMonkey
{
    private const string legalCharacters = " abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ1234567890.";
 
    static Random random = new Random();
 
    /// The Typing Monkey Generates a random string with the given length.
    public string TypeAway(int size)
    {
        StringBuilder builder = new StringBuilder();
        char ch;
 
        for (int i = 0; i < size; i++)
        {
            ch = legalCharacters[random.Next(0, legalCharacters.Length)];
            builder.Append(ch);
        }
 
        return builder.ToString();
    }
}

You can find the source code here: http://github.com/JohnIdol/typingmonkey/tree/master

Feel free to contribute or despise.