General > Biology

Costs and energy usage.

<< < (5/5)

EricL:
Personally, I'm not a big fan of genotipic costs - that is, costs associated with DNA operators - or with costs that are associated with meta aspects of the DNA such as it's length or the number of bps executed, etc.   Such costs provide selective pressure, but towards minimalism I.e. shorter genomes or the execution of the miunimal number of base pairs or minimal use of specific base pairs.  

I much prefer morphological and environmental costs - that is, costs that are associated with what the bot actually is or does, not on the code it uses to do it.   This subject touches on the separation between genotype and pheneotype, something that is half pregnant in DB.  In biology, having more DNA is basically free.  Even 'executing' more DNA is basically free given the natrual world's inherent parallelism.  What matters is what the end result is.  Larger bodies/beaks/penises has a real cost in the real world.  What specific DNA was executed to create those larger brains/trunks/fins can bascially be ignored as far as costs go.  Inheritability and evolvability though, well, those do influence DNA structure and exhert selection pressures, but that is different from what IMHO should be purely morphological costs...

Testlund:
Well, I agree about that it's just that I've been thinking that the dna command costs could be the same as bodily functions in a living cell. Maybe I should set all DNA Command Costs to 0 but maybe have a small cost on Flow Command to simulate bodily function costs in a living cell.

Maybe you could change the default values for the next version so all I need to do is to click on the Default button and the values that makes most sense whould be put in there.  

Numsgil:
The end thing really is that whatever the costs are, the bots are going to adapt against it.  If you aren't really sure what your trying to find (that is, you're just interested in finding something interesting) just set the various costs to different things.

Just be sure that you'll be able to figure out what and why a bot is doing things.  If you can invision a root cause for every adaptation you'll see in the sim you'll slowly start understanding what the various costs are doing and how to set them.

Testlund:
I think that the best way to approach this problem is to just put a cost in one field at a time and then compare how it will affect a bot through a certain amount of cycles, and then find a balance between day/night cycles and how the bot spend it's energy, maybe that's what you mean, nums. I did an experiment and found that some costs doesn't work at all! Flow command costs for instance does nothing! So with day/night cycles set to 10000 I found these costs to work best:

Translational Bang: 0.23
Tie Formation: 5 (just put 5 here without testing it)
DNA upkeep: 0
Body upkeep: 0.000025

The other morphological costs I set to 1.

All DNA command costs I set to 0.

Age cost: 0.1
Begins upon reaching 20000 cycles.
Increase cost log(age)

I did an experiment with putting just 1 Animal Minimalis and 1 R Fisannis in the sim to see how long they whould survive. R Fisannis can live for about 10000 cycles swimming around constantly while Animal Minimalis which was sitting still survived for about 20000 cycles. There was no food for them. A veggie on the other hand wont lose any energy at all while the sun is up, but will lose a certain amount when the sun is down based on it's size. The larger the more it will lose each cycles, but on the other hand it will also gain energy faster because of it's size when the sun is up. Aint that cool?.  

Numsgil:
Yep, that's more or less what I was suggesting.  You just have to sort of get your hands dirty to figure things out.

Navigation

[0] Message Index

[*] Previous page