Page 3 of 5

Re: Randomized Swimmer Color and Pseudo Races

Posted: Fri Jun 23, 2017 8:30 pm
by bwisialo
I'm able to go from a Stereocyte egg dorectly to the two Stemocytes for four color combinations, saving two modes. But B,O and O,B are extremely rare, like 1 out very 1000 or something.

Re: Randomized Swimmer Color and Pseudo Races

Posted: Fri Jun 23, 2017 9:14 pm
by C4pung
Wow the dumb version was actually alot more complex than I thought! :shock: . I wrote the design of that kite swimmer so the modes are clear and efficient. 2 colors on 4 cells would mean 2^4 (16) combinations total right?.

I must try out the linear one to see if its possible, kite version was really close, it took 19 modes. I dont think its possible to have 4 x 4 color combination but we might just be able to make 4 x 2 or 4 x 1 color using a modified stereocyte 90/10 split. So we stacked stereo --> stemo --> stereo --> stemo --> color choice. The second stereocyte should be angled diffrently so it will have diffrent results than the stereocyte before.

Re: Randomized Swimmer Color and Pseudo Races

Posted: Sat Jun 24, 2017 7:21 am
by C4pung
The linear version wasnt really hard to make but optimizing it was kinda a trouble, they wiggle alot :?

Subs : https://www.dropbox.com/s/ql5xo58euj1hq ... trate?dl=0

Image

Like before the eggs split into 4 stereocytes that detects food, then 90/10 split into a stemocyte that determines the color of their body parts. If no food is detected the stemocyte will 90/10 split back to the stereocytes until they found a food. This time I played around the child angles so I can change the stereocyte's antennae without comprimising the layout.

I used Wapcaplet's splitting linear swimmer color layout :P . The modes and color codes :
M12 : Flagello color choice 1 (pink)
M13 : Flagello color choice 2 (purple)
M14 : Myo color choice 1 (pink)
M15 : Myo color choice 2 (purple)
M15 : Stereo color choice 1 (pink)
M16 : Stereo color choice 2 (purple)
M17 : Phago color choice 1 (pink)
M18 : Phago color choice 2 (purple)

I have an idea, how about a stealth / chameleon type swimmer that mimics its predator's color? . Also is it possible to have 3 color instead of even color?

Re: Randomized Swimmer Color and Pseudo Races

Posted: Sat Jun 24, 2017 7:35 am
by bwisialo
You got all four randomized! Nice! :D

You can do three colors. It doesn't even necessarily require more Stereocytes, though it does require an additional set of Strmocytes. It can be as simple as this. Stemocyte 1: if c>.02, then Red; if c <.02, then Stemocyte 2. Stemocyte 2: if c < -.02, then Green; if c> -.02, then blue. In short, strong positive is red, strong negative is green, and weak or no signal is blue. Probably difficult to get the probability equal to one third for each.

A chameleon with two color options should be relatively easy -- easier than randomizing all cells. It would be more like: if green is detected then go to egg for green organism; if red is detected....

Re: Randomized Swimmer Color and Pseudo Races

Posted: Sat Jun 24, 2017 8:14 am
by C4pung
Thanks :D . It took the same exact amount of modes :lol: . The only things that were diffrent was the cell types and splits but the mechanism is the same.

I would really like to make a habitat with both color changing predators and prey!. Prey are able to change their offapring colors based on the presence of predators, and predators make their offspring based on what prey is available to hunt. It will take alot of optimization and a huge plate to reach equilibrium though :( .

For prey I think the Pseudo grouping is sufficient, I need to try your stemocyte configs and see how it works out. For predator I need to make a new one with 2 senseocytes :? , is it possible to just use a traditional layout but use a diffrent mechanism like this :

Reproductive organ make 1st senseo detecting 1 color --> 90/10 split becoming another senseocyte detecting the other color --> 90/10 split to a stemocyte --> stemocyte channels to the stronger color signal

Re: Randomized Swimmer Color and Pseudo Races

Posted: Sat Jun 24, 2017 8:51 am
by bwisialo
I don't think that predator /prey set up should be all that difficult. Then again, I recall trying some version of that and the direct route that I thought would work didn't pan out right away. :lol: But I think I just got the differentiations wrong. Equilibrium might be a different story.

I'm not sure pseudo grouping will happen here. Eg, red predator seeks green prey; when the red predator is near, green prey will not produce more green prey but will produce red prey. The same will happen with red prey producing green prey when the green predator is around. Predators will have to be near if they are going to survive. But maybe I'm imagining this differently than you.

I don't think you want one Senseocyte, then a second, then Stemocyte. I think that would give the second Senseocyte a significant advantage as substance from the first one decreases. You could have the Senseocytes in the swimmer itself, signalling to a Stemocyte egg. Or, you can have a Lipocyte split into Stemocyte and Lipocyte, and the second Lipocyte splits into two Senseocytes for the two colors.

Daaang! Complicated stuff. :lol:

Re: Randomized Swimmer Color and Pseudo Races

Posted: Mon Jun 26, 2017 1:45 am
by C4pung
Here is a prototype for the chameleon type swimmer, pretty easy but also kinda plain :
https://www.dropbox.com/s/sfynmh9z9q1vg ... trate?dl=0

Cyan swimmers are the "real prey", predator hunts them. Both of the cyan and white "ghost" swimmer produce a senseocyte that will detect predator. After that it will 90/10 split into a stemocyte, the stemocyte will make cyan swimmers if there are no predators, and make ghost ones if there are predators around.

I its possible to make smart prey version of this (using kite instead traditional to reserve modes), but I think it will be too OP for the predator. Any suggestion?

Re: Randomized Swimmer Color and Pseudo Races

Posted: Mon Jun 26, 2017 4:28 am
by bwisialo
Yeah, I've wondered if camouflage would mean predators just die. I'm not sure. It will be interesting to find out. I'm assume it can be made to work one way or another. You will want a larger substrate to do this, maybe 2-3mm.

If the predators do die because of color camouflage, you could try what we mentioned before: two predators in the same genome; one hunts cyan, the other hints white, depending on which one is detected at birth.

Did you mean kite to save modes? Kite smart prey and other smart prey use the same number modes, nine, so no modes are saved.

If you want to save modes, you could, eg, give the cyan swimmer the same white Flagellocytes as the ghost: the whole swimmer doesn't need to be cyan.

Re: Randomized Swimmer Color and Pseudo Races

Posted: Mon Jun 26, 2017 4:52 am
by C4pung
Ah right, dumb of me :lol: . I kinda work better at kites though, less swaying with minimal programming. Come to think of it, about 50% of my swimmers I posted lately were based on kite swimmers.

For the predator that would be kinda tough for me using the same stemocyte but I will try. What do you suggest for the predator layout? Easiest one would be super-smart predator layout but that way I need to put the stereocyte way back, that will hinders it movements :? . Im going to try your Brain predator layout.

Re: Randomized Swimmer Color and Pseudo Races

Posted: Mon Jun 26, 2017 5:12 am
by bwisialo
In my experience, kite smart prey are much floppier and sway a lot more than the traditional, because of the hole in the middle. Why is programming the traditional more difficult?

I meant, use a traditional predator layout. Does the predator need two Stereocytes? Super-smart has Stereocytes in the middle, not rear. I'm confused. :? :lol:

Rear Stereocyte doesn't hinder movement at all.