Welcome To Darwinbots > Newbie
AZPaul Intro
AZPaul:
Howdy y'all.
I found Darwinbots v2.3 a few days ago and am getting used to the function. Nice program. I have coded up a small number of Bots that actually live, feed, reproduce and like that. At this point I am running some simulations but am having some difficulties with a major item. I'm hoping y'all can set me right.
Sexual reproduction.
I use the .in/.out as species ID. According to the docs the genome of the "closest body" is taken as the sex partner. I assumed (silly me) that I could ID this "closest body" using .in/.out. It don't work. Appears the .in is populated from the last bot I "saw" which, in too many cases, is not the closest body when .nrg reaches the threshold set for my .sexrepro gene. Right so far?
Mutation in allele frequency is the source of almost all genetic change (most being deadly), but sex (shuffeling of complete genes from those that have survived their mutations to form a unique genome) is the prime vector of "survivable" variation within a large speciel population. OK, so a couple hundred Bots is not really a "large" speciel population, but, we deal with what we got, right?
For my simulations I cannot have my Alpha mating with some wayward Delta. Especially a mean nasty Delta that feeds on Alphas. Poisons the gene pool doncha know and that's just bad form. Mating outside the species is a no-go in our version of real life and cannot be allowed in the simulation.
This is one of the most powerful features within NeoDarwinian Evolution and that's where I want to be.
Any of y'all have a workaround?
Cosmetic items:
.backshot -- Doesn't want to work for me in the way I have it applied. Bots discharge waste. I load up .shootval then -4 .backshot. Nothing happens. Waste continues to accumulate. I substitute -4 .shoot in the gene with no other changes and the Bot pukes up the waste just fine. Don't ask why, but I want to make them poop not puke.
If there is a .valbackshot or .backshotval or whatever I cannot find it in the docs. Maybe I cannot .backshot through a shell?
Any help?
Graphs -- I run my simulations for about 3-5 hours each. (The "best Bot" save feature (I set for 30 min) is much appreciated.) I like to have the population graph up on my desktop so I can keep track that nothing drastic is happening, like everyone is dying. Problem is that while I'm in my other applications doing some "real work" this graph is alway "on top" and in my way. I will look at it when I want to on the desktop and don't need this thing in the way of my spreadsheets, web pages and documents. Are there any options to "alway on top?"
Finally, (I promise) I diddle around some, changing this, that, the other option, resetting graphs, starting, stopping, setting up simulations and etc like that. I've managed to generate "run time errors" frequently. Considering what I paid for this program (another much appreciated feature) this is just an FYI.
OK, I lied. One more comment. I appreciate the efforts you folks put into this program. I know a lot of you people are building MonsterBots for competition, but as a learning tool in NeoDarwinian Evolution and simulating atleast some aspects of natural selection this program is already one of the better feature-rich simulators without undue complexity. And I thank you.
Verbosely yours,
-P
Botsareus:
Ok first of all were did you get the name AZPaul does it mean "alpha zone paul"?
Do you know a player that goes by the name TerminatorDelta ?
--- Quote ---I use the .in/.out as species ID. According to the docs the genome of the "closest body" is taken as the sex partner. I assumed (silly me) that I could ID this "closest body" using .in/.out. It don't work. Appears the .in is populated from the last bot I "saw" which, in too many cases, is not the closest body when .nrg reaches the threshold set for my .sexrepro gene. Right so far?
--- End quote ---
In works from the eye subrotine , it is not Necessarily the closest body. You can have a bot looking on a bot far away while: there is bot behind your bot witch
is very close, so it will reproduce based on that bot.
But I dont know for sure , PY and NUM studed the source code, they should be able to tell if thats a bug or not.
P.S.
--- Quote ---Cosmetic
--- End quote ---
Good referance to Cosmic Rift
--- Quote ---For my simulations I cannot have my Alpha mating with some wayward Delta. Especially a mean nasty Delta that feeds on Alphas. Poisons the gene pool doncha know and that's just bad form.
--- End quote ---
Talk about alpha zone , and me delta....
Give it up....
shvarz:
Welcome AZ!
I'll answer what I can (which is not much):
1. Agree on sexrepro. It should be changed for cospecifics only. I don't think it allows crossing to veggies though.
2. No clue about backshot, did not know such thing existed.
3. Graphs - just get used to it :) I did. Ask Nums or PY to change that.
4. Run time errors are frequent, and Nums and PY are good at fixing them if you can tell them what exactly you do that is causing it.
5. If you want to run evolution sims, then you came to the right place. DBs is great for it! That's what I mostly do, so you are not going to be alone!
Botsareus:
--- Quote ---Mating outside the species is a no-go in our version of real life and cannot be allowed in the simulation.
--- End quote ---
Ok we dont need to change the program to do this, Simply check some .refeye and somre .in .in2 and stuff check all the $%$ .ref stuff , Using them , build a robot that chooses witch "species" to sexrepro to. If you need help on building robots thats another story.
PurpleYouko:
Welcome to DarwinBots.
Here are a couple of tips
To use .sexrepro with the correct partner why not first find the right partner then tie to it to make sure it stays nice and close.
next use *.multi to make sure the tie is good and hard then rotate the tie using .tienum and .setang. this makes sure you have an open space in front.
Now you can use .sexrepro to mix genes between your two tied bots with the reasonable certainty that they are thee physically closest couple.
Next use .deltie to delete the tie and be off on your way.
Truthfully though .sexrepro is not the most successful inovation that has ever been added. We don't use it too much at all really.
As for .backshot, all you have to do is put a 1 in it to enable back-shooting for that cycle. It resets to zero for the next cycle. Then just fire the waste shot as normal using -4 .shoot store. The waste will come out the back. Use a gene like this..
--- Code: ---cond
*.waste 100 >
start
1 .backshot store
-4 .shoot store
*.waste .shootval store
stop
--- End code ---
I never use the graph functions. In my opinion they don't work properly, never did work properly and are next to useless. I didn't program them and have little interest in re-programming them either. They are a superfluous and useless feature that I would be more than happy to ditch completely.
Tell us your problems and bugs and they will be addressed. Maybe not in V3.0 but sooner or later.
:D PY :D
Navigation
[0] Message Index
[#] Next page
Go to full version