Also the software package that Paul and I have been working on is really starting to work great. My half is still a little glitchy, consuming system memory a bit more than I'd like but I can now test and breed around 10,000 fractals before anything goes haywire. Saturday night was the big breakthrough when I implemented a tournament breeding routine which delivers some very accurate approximations of the best fractals compared to the target image. Here's my first really cool image:


Not a bad match I must say. Only took about 12 generations to arrive at that. That's 100 individual fractals per generation with the tournament breeding routine in between. What do you know, I can even farm mathematics.
1 comment:
Very nice! I still haven't quite figured out the fractal stuff.
Post a Comment