Cellular Automa - comp.programming

I had several ideas for cellular automa that I was interested in trying out, but I would like any suggestions. I am currently just laying out the basic system of the area that the automaton will be held in. I have two types specifically in my mind: Traditional automaton in which each cell has two states; dead and alive. I would have more time to work and make area and cell placing systems more advanced, but it just seems like so many people have already done this... Also, there would be the energy/strength model. It would run along the same system of the original, except that it will be based upon an energy or strength system. For example: if(cell[x][y]>5||proximity(x,y)==0){cell/2;cell[x+1][y+1]==cell[x][y]*2} Also, say the cell is surrounded by 8 others; if it has a greater energy than those around it, it will gain one energy. If its energy is equal to the greatest of its peers, it will survive. If it is less than its peers, it will die. Just some ideas, thanks for any input.

Flaran said: > I have two types specifically in my mind: > > Traditional automaton in which each cell has two states; dead and > alive. I would have more time to work and make area and cell placing > systems more advanced, but it just seems like so many people have > already done this... Yeah, it's Game of Life, basically. You can tweak the rules, but it's still Game of Life under the hood. > Also, there would be the energy/strength model. It would run along the > same system of the original, except that it will be based upon an > energy or strength system. > For example: > > if(cell[x][y]>5||proximity(x,y)==0){cell/2;cell[x+1][y+1]==cell[x][y]*2} > > Also, say the cell is surrounded by 8 others; if it has a greater > energy than those around it, it will gain one energy. Conservation of energy - where does this extra energy come from? > If its energy is > equal to the greatest of its peers, it will survive. If it is less > than its peers, it will die. This is a bit more interesting. I personally think it's more interesting to model "creatures" which have freedom of movement around the grid. But if you want to stick to static "creatures", consider the possibilities inherent in having more than one species. For example, a cell might be empty, but surrounded by, say, three or four "grass" cells, in which case this cell becomes "grass". A "grass" cell surrounded by too many "grazer" cells might die out. A "grazer" cell adjacent to a "growler" cell could be terminally affected by the experience. A "growler" cell without sufficient "grass" around it might die from hunger because it has insufficient camouflage. That could provide some interesting dynamics, even with static "creatures". IMHO getting them to move around gives a lot more scope, but then it's not really a cellular automaton any more. -- Richard Heathfield "Usenet is a strange place" - dmr 29/7/2005 http://www.cpax.org.uk email: rjh at above domain (but drop the www, obviously)

On Wed, 26 Oct 2005, Richard Heathfield wrote: > Flaran said: >> >> Traditional automaton in which each cell has two states; dead and >> alive. I would have more time to work and make area and cell placing >> systems more advanced, but it just seems like so many people have >> already done this... > > Yeah, it's Game of Life, basically. You can tweak the rules, but it's still > Game of Life under the hood. Depends on what you mean by "Game of Life." How about Stick Wars 664? http://www.collidoscope.com/modernca/stickwarsrules.html Is that still "Game of Life under the hood"? It certainly bears little resemblance to Conway's game. "Maze" and "Day and Night" (under "Life-Like Patterns" on the same site) are also interesting, and actually even /less/ like regular Life. Multi-state automata are fun to experiment with. A cellular-automaton engine is a good small project for a beginner-to-intermediate programmer, I'd say. (I made one!) >> Also, there would be the energy/strength model. It would run along the >> same system of the original, except that it will be based upon an >> energy or strength system. >> For example: >> >> if(cell[x][y]>5||proximity(x,y)==0){cell/2;cell[x+1][y+1]==cell[x][y]*2} >> >> Also, say the cell is surrounded by 8 others; if it has a greater >> energy than those around it, it will gain one energy. > > Conservation of energy - where does this extra energy come from? Who says energy has to be conserved? >> If its energy is >> equal to the greatest of its peers, it will survive. If it is less >> than its peers, it will die. > > This is a bit more interesting. Really? Seems to me as if a lot of cells die, and no new cells are created --- thus leading to a steady state in approximately three turns. I'd call that a really boring CA. > I personally think it's more interesting to model "creatures" which have > freedom of movement around the grid. But if you want to stick to static > "creatures", consider the possibilities inherent in having more than one > species. For example, a cell might be empty, but surrounded by, say, three > or four "grass" cells, in which case this cell becomes "grass". A "grass" > cell surrounded by too many "grazer" cells might die out. A "grazer" cell > adjacent to a "growler" cell could be terminally affected by the > experience. A "growler" cell without sufficient "grass" around it might die > from hunger because it has insufficient camouflage. > > That could provide some interesting dynamics, even with static "creatures". Sounds like one way of thinking about multi-state automata, albeit a way rather hobbled by its real-world metaphor. > IMHO getting them to move around gives a lot more scope, but then it's not > really a cellular automaton any more. Yep. But multi-state CAs can have moving creatures, as long as you define "creatures" more abstractly --- like the inchwormy things in "Worms" (25/3467,6) or the spaceships in "Wanderers" (34678/345,5). Or the gliders, spaceships, puffers, etc., in Conway's Game of Life (3/23), for that matter. Of course it's a lot harder to get CAs to do any particular interesting thing (model predator-prey dynamics, move in a circular trajectory, etc.), but once you start looking at multi-state automata, you find a lot of really interesting-looking things they do on their own. The OP might want to check out comp.theory.cellular-automata. -Arthur

Thanks for the input, and I had thought about trying out something like that; the only problem is how to get it to move around the grid. I'm not really sure exactly what would be the best way to go about that. Maybe have a 20 by 20 or so array; if the value is 0 then it is a grass cell. If the value is one, it is a grazer. If the value is two, it is a growler. When the grazer/growler moves... grid[x+1][y]=grid[x][y];grid[x][y]=0;x+1;

In article <1130355385.372792.179990@g43g2000cwa.googlegroups.com>, flaran@gmail.com says... > Thanks for the input, and I had thought about trying out something like > that; the only problem is how to get it to move around the grid. I'm > not really sure exactly what would be the best way to go about that. > > Maybe have a 20 by 20 or so array; if the value is 0 then it is a grass > cell. If the value is one, it is a grazer. If the value is two, it is > a growler. This is 2005. I would suggest an array of 200 x 200 minimum. - Gerry Quinn

In article <4360ab6c@news.orcon.net.nz>, not@interesting.com says... > Gerry Quinn wrote: > > > This is 2005. I would suggest an array of 200 x 200 minimum. > > This will look funny in ten years. Probably! My thinking, though, is that there is likely to be a lot of interesting activity in between the 20-200 range for typical projects of this size. 20x20 will be cramped for almost anything you want to do - 200x200 will be a lot less so for many things. And it's a nice size for watching the lot on a typical monitor. - Gerry Quinn