Darwinbots Forum

Code center => Suggestions => Topic started by: Endy on May 06, 2005, 04:40:38 AM

Title: Mutation Suggestion
Post by: Endy on May 06, 2005, 04:40:38 AM

If it is best for evolution for the children to be only 1% or less different than their parents would it be possible to simply count the bot's dna and let the program mutate only 1% or less of the child's dna?

Not sure how hard this would be to implement, but it should make guestimating the best Mutation ranges easier.

Endy B)

Title: Mutation Suggestion
Post by: Numsgil on May 06, 2005, 01:53:15 PM

My only criticism would be that it's not very realistic.  Other than that it shouldn't be too hard.

Title: Mutation Suggestion
Post by: Zelos on May 06, 2005, 02:10:13 PM

there shouldnt be a limit, but the there is a mutationer/X basepairs/generation. I dont remember what X is, but it had a lower value long time ago. the more dna the more mutation there is

Title: Mutation Suggestion
Post by: shvarz on May 06, 2005, 02:10:22 PM

Yes, that is essentially what I was asking for.  It is not very realistic, that is why I asked for a tool that would help calculate that.

Title: Mutation Suggestion
Post by: Numsgil on May 06, 2005, 02:14:04 PM

That's what I'm thinking, a quick built in tool to calculate it wouldn't be hard.  Consider it on the list, lol.

Title: Mutation Suggestion
Post by: Endy on May 07, 2005, 01:57:24 AM

Yeah, not too realistic, but I get tired of either having too many or too little mutations. Too little and the sim is boring or quickly ended in harsh conditions; too much and every bot is almost completly unrelated to one another.

Thanks for putting it on the List :)

Endy B)