11,643,493 members (73,719 online)

# AI: Dawkins Biomorphs / And Other Evolving Creatures

, 29 Jan 2007 CPOL 101.5K 525 53
 Rate this:
An implementation of Dawkins Biomorphs and some more crazy evolving creatures.

## Introduction

A while ago, I found something entitled Biomorphs, which appeared to be a bunch of random drawings created by some sort of computer program. I began investigating this a bit further, and found out this was actually the result of running a popular AI type program, which is carrying out an algorithm described by Richard Dawkins from his "The Blind Watchman" book, entitled BIOMORPHS.

Dawkins investigated the possibilities inherent in just the combination of mutation and selection, when combined with a very powerful development (decoding) process.

This is basically a recursive algorithm that has some genes and uses some mutation and some random selection. That's it. That is all that is required to produce a wide variety of fascinating virtual creatures, some which look very genetic.

## Dawkins Basic System

The phenotypes (how it looks) in Dawkins system were essentially branching tree structures. An extension is to add segments to this (this is not included in this article).

The basic system has nine genes, controlling things like:

• angle of branching
• depth of branching
• number of lines
• etc. etc.

It makes use of a recursive algorithm to carry out the drawing; this is shown below:

```private void draw(Graphics g, int i, int j)
{
tree(g, i / 2, j, order, 2);
}

/// <summary>
/// Draws the Dawkins bio-morph structure (No segmentation, in this
/// implementation)
/// </summary>
/// <param name="g">The graphics to use</param>
/// <param name="i">The x position to start drawing</param>
/// <param name="j">The x position to stop drawing</param>
/// <param name="k">The y position to start drawing</param>
/// <param name="l">The y position to stop drawing</param>
private void tree(Graphics g, int i, int j, int k, int l)
{
try
{
if (l < 0)
l += 8;
if (l >= 8)
l -= 8;
int i1 = i + k * dx[l];
int j1 = j - k * dy[l];

g.DrawLine(new Pen(new SolidBrush(Color.White), 1.8F), i, j, i1, j1);
if (k > 0)
{
tree(g, i1, j1, k - 1, l - 1);
tree(g, i1, j1, k - 1, l + 1);
}
}
catch (Exception)
{
}
}```

Using this algorithm, it is possible to come up with some really interesting creatures such as:

They look very organic, don't they? However, if we introduce a segmentation (not done for this article) to the algorithm, we can then achieve some truly bizarre creatures such as these. Basically, we are using the same idea, but we are creating more segments of the same basic shape.

## Class Design

The following diagram illustrates the code design. I will give a brief discussion on each class before I show the demonstration screenshots.

Basically, there are two types of bio-morphs that I created for this article. These are:

• BioMorph, which makes use the the following classes:
• `BiomorphPanel`: Which provides the rendering for the original Dawkins Biomorph.
• `IBioEvolvalbe`: An interface which is implemented by the `BioMorphPanel` class.
• `BiomorphLayoutPanel`: Hosts a number of `BiomorphPanel` objects. This is then hosted on a Windows Form, and it is responsible for the selection of a random `BiomorphPanel`, to spawn a new generation of Dawkins Biomorphs.
• EvolvingCritter, which makes use the the following classes:
• `EvolvingCritterPanel`: Which provides the rendering for a Cow Skull type Biomorph.
• `IEvolvalbe`: An interface which is implemented by the `EvolvingCritterPanel` class.
• `CritterLayoutPanel`: Hosts a number of `EvolvingCritterPanel` objects. This is then hosted on a Windows Form, and it is responsible for the selection of the fittest `EvolvingCritterPanel`, to spawn a new generation of Dawkins Biomorphs.
• `EvolvingCritterComparator`: Provides a simple implementation of `IComparer`, to allow a collection of `EvolvingCritterPanel` objects to be sorted in order to obtain the fittest one.

There is also a common interface:

• `ILayoutPanel`: which is implemented by both `BiomorphLayoutPanel` and `CritterLayoutPanel`, which allows the Windows Form where one of these two panels is hosted to generate a new population of organisms.

## Screenshots of the Running Application

OK, so what does the application look like? Well, it depends on which of the Biomorphs is currently in use. Basically, there are two links, and depending on which one is selected, the main panel will be swapped to show the correct Biomorphs panel. It will either be a `BiomorphLayoutPanel` (Dawkins Biomorphs), or a `CritterLayoutPanel` (my own Cow Skull Biomorphs). Let's have a look at this, shall we?

### Using Dawkins BiomorphLayoutPanel BioMorphs

There is a timer placed on the form which, every 1500 milliseconds, will choose a random Dawkins Biomorph, and will use that to spawn a new population of crazy critters. There is only random selection (no tricks), and you can get some very bio-logical looking critters. Basically, it's all down to the clever gene decoding we saw earlier.

### Using CritterLayoutPanel BioMorphs (My Own Cow Skull Design)

I really only included the possibility to create other style Biomorphs for a bit of fun. The difference here is that these chaps are judged to see which one should be the one that is used for breeding. Basically, the one with the most area, and the preferred eye color, and the preferred constituent parts, will have a better score, so it will more likely be the alpha male that will be used for breeding.

Note: We do actually pick the fittest one here, so one would hope that over time, we would end up with similar shaped or colored organisms. But nature has other ideas, mutation is very powerful. Play with it a while and you will see.

### Freedom of Choice

With both these Biomorph panels, the user is also free to click on the organism they like; then this will be used as the new alpha male that is then used to create a new population.

## What Do You Think?

That's it. I would just like to ask, if you liked the article, please vote for it.

## Conclusion

I have quite enjoyed constructing this article. I hope you liked it.

## History

• v1.0: 29/01/07.

## Share

 Software Developer (Senior) United Kingdom
I currently hold the following qualifications (amongst others, I also studied Music Technology and Electronics, for my sins)

- MSc (Passed with distinctions), in Information Technology for E-Commerce
- BSc Hons (1st class) in Computer Science & Artificial Intelligence

Both of these at Sussex University UK.

Award(s)

I am lucky enough to have won a few awards for Zany Crazy code articles over the years

• Microsoft C# MVP 2015
• Codeproject MVP 2015
• Microsoft C# MVP 2014
• Codeproject MVP 2014
• Microsoft C# MVP 2013
• Codeproject MVP 2013
• Microsoft C# MVP 2012
• Codeproject MVP 2012
• Microsoft C# MVP 2011
• Codeproject MVP 2011
• Microsoft C# MVP 2010
• Codeproject MVP 2010
• Microsoft C# MVP 2009
• Codeproject MVP 2009
• Microsoft C# MVP 2008
• Codeproject MVP 2008
• And numerous codeproject awards which you can see over at my blog

## You may also be interested in...

 First Prev Next
 My vote of 5 Kenneth Haugland16-Sep-13 3:20 Kenneth Haugland 16-Sep-13 3:20
 changing code Kitten8310-Dec-11 16:31 Kitten83 10-Dec-11 16:31
 Re: changing code Sacha Barber10-Dec-11 21:38 Sacha Barber 10-Dec-11 21:38
 Leaving Java Behind logicchild9-Mar-09 22:39 logicchild 9-Mar-09 22:39
 [Message Deleted] Danny Rodriguez27-Jan-08 9:11 Danny Rodriguez 27-Jan-08 9:11
 Good Stuff! bobsugar2221-Feb-07 4:52 bobsugar222 1-Feb-07 4:52
 Re: Good Stuff! Sacha Barber1-Feb-07 7:41 Sacha Barber 1-Feb-07 7:41
 Antithesis Ilíon30-Jan-07 6:17 Ilíon 30-Jan-07 6:17
 Re: Antithesis Sacha Barber30-Jan-07 6:53 Sacha Barber 30-Jan-07 6:53
 Re: Antithesis Ilíon30-Jan-07 7:24 Ilíon 30-Jan-07 7:24
 Re: Antithesis [modified] Sacha Barber30-Jan-07 8:00 Sacha Barber 30-Jan-07 8:00
 Re: Antithesis Ilíon30-Jan-07 8:53 Ilíon 30-Jan-07 8:53
 To further elucidate this antithetic discrepancy between algorithms and Dawkins' message, the point at issue is 'information.' Dawkins teaches (and his world-view requires it to be true) that 'information' comes from no-where, that it just poofs into existence -- it's all so very mysterious, really -- unbidden; that 'information' is self-created or self-generated. And he presents 'biomorphs' as somehow proving his claims. Now, I realize that there exist programmers who, for whatever reason(s), may subscribe to such a view. But really, we programmers, of all people on this planet, simply have no excuse for swallowing such a patent falsehood: *we* create 'information.' I don't, of course, mean that we programmers and we alone create 'information,' but rather that the creation and manipulation of 'information' is the explicit and specific task we engage in when we program. We (as individuals) may not understand *everything* about 'information,' but surely we ought to understand some basic things about it; starting with: it does not generate itself from nothing, but rather is always and only produced by a 'mind.' You have an algorithm you've created; you have some set or sets of numbers to use as inputs to that algorithm. You feed the inputs to the algorithm, and mirible dictu, the algorithm generates interesting pictures. But there was nothing "magical" about this; no 'information' poofed into existence when you exectued the algorithm. Rather, the 'information' was already fully contained within the algorithm you wrote. The 'information' is not the inputs -- which are, in themselves and by themselves, meaningless numbers -- it is in the algorithm, the pre-written set of steps for ascribing some certain meaning to a set of inherently meaningless numbers. Your algorithm may well generate all sorts of vaguely interesting pictures given all sorts of non-deterministic sets of inputs; for that is what it is written to do. The 'information' for generating those pictures is already fully contained in the algorithm. But, it will never, ever, produce, for instance, the schematics for the Empire State Building. That particular 'information' is not contained in that particular algorithm.
 Re: Antithesis Sacha Barber30-Jan-07 9:06 Sacha Barber 30-Jan-07 9:06
 Re: Antithesis Ilíon31-Jan-07 3:47 Ilíon 31-Jan-07 3:47
 Re: Antithesis Sacha Barber31-Jan-07 4:14 Sacha Barber 31-Jan-07 4:14
 Re: Antithesis un_ko31-Jan-07 2:38 un_ko 31-Jan-07 2:38
 Re: Antithesis Sacha Barber31-Jan-07 4:07 Sacha Barber 31-Jan-07 4:07
 Re: Antithesis un_ko31-Jan-07 11:57 un_ko 31-Jan-07 11:57
 Re: Antithesis Sacha Barber1-Feb-07 7:43 Sacha Barber 1-Feb-07 7:43
 Re: Antithesis Sacha Barber1-Feb-07 7:44 Sacha Barber 1-Feb-07 7:44
 Re: Antithesis [echo]6-Feb-07 20:43 [echo] 6-Feb-07 20:43
 Re: Antithesis Sacha Barber6-Feb-07 22:25 Sacha Barber 6-Feb-07 22:25
 Re: Antithesis M. Thomas Frederiksen25-Dec-07 18:11 M. Thomas Frederiksen 25-Dec-07 18:11
 Re: Antithesis Lawrence Botley8-Feb-07 2:51 Lawrence Botley 8-Feb-07 2:51
 Re: Antithesis Ilíon8-Feb-07 4:50 Ilíon 8-Feb-07 4:50
 Re: Antithesis Lawrence Botley8-Feb-07 5:19 Lawrence Botley 8-Feb-07 5:19
 Re: Antithesis Sacha Barber8-Feb-07 5:54 Sacha Barber 8-Feb-07 5:54
 Re: Antithesis [modified] Ilíon8-Feb-07 6:05 Ilíon 8-Feb-07 6:05
 Re: Antithesis Sacha Barber8-Feb-07 7:33 Sacha Barber 8-Feb-07 7:33
 Re: Antithesis Ilíon8-Feb-07 7:44 Ilíon 8-Feb-07 7:44
 Re: Antithesis Sacha Barber8-Feb-07 8:15 Sacha Barber 8-Feb-07 8:15
 Re: Antithesis easyTree4-Mar-07 3:18 easyTree 4-Mar-07 3:18
 Re: Antithesis mhunt1311-Feb-07 23:21 mhunt13 11-Feb-07 23:21
 Re: Antithesis Sacha Barber12-Feb-07 0:15 Sacha Barber 12-Feb-07 0:15
 Re: Antithesis Ilíon12-Feb-07 3:12 Ilíon 12-Feb-07 3:12
 Re: Antithesis Sacha Barber12-Feb-07 3:39 Sacha Barber 12-Feb-07 3:39
 Re: Antithesis Ilíon12-Feb-07 3:59 Ilíon 12-Feb-07 3:59
 Cool... Peter Hancock29-Jan-07 11:54 Peter Hancock 29-Jan-07 11:54
 Re: Cool... Sacha Barber29-Jan-07 21:23 Sacha Barber 29-Jan-07 21:23
 Interesting stuff WillemM29-Jan-07 9:45 WillemM 29-Jan-07 9:45
 Re: Interesting stuff Sacha Barber29-Jan-07 11:10 Sacha Barber 29-Jan-07 11:10
 Interesting Article aprenot29-Jan-07 9:03 aprenot 29-Jan-07 9:03
 Re: Interesting Article Sacha Barber29-Jan-07 9:13 Sacha Barber 29-Jan-07 9:13
 Re: Interesting Article aprenot29-Jan-07 9:31 aprenot 29-Jan-07 9:31
 Re: Interesting Article Sacha Barber29-Jan-07 11:09 Sacha Barber 29-Jan-07 11:09
 Re: Interesting Article aprenot29-Jan-07 11:50 aprenot 29-Jan-07 11:50
 Re: Interesting Article Sacha Barber29-Jan-07 21:10 Sacha Barber 29-Jan-07 21:10
 Last Visit: 31-Dec-99 18:00     Last Update: 2-Aug-15 22:15 Refresh 1