[Tiled CA program] - A New Kind of Science: The NKS Forum

A New Kind of Science: The NKS Forum

Pages:1


Tiled CA program

(Click here to view the original thread with full colors/images)


Posted by: edpegg

A wonderful CA program - Tiled CA has been made by Brian Prentice. It can be related to page 930 - Other Geometries.

Basically, the program allows you to create a tiling, and then to build CA rules for that tiling. You may also just generate rules randomly. The program comes with about a hundred different tilings to look at, as well as many rules.

http://linuxenvy.com/bprentice/TiledCA/TiledCA.html

Brian welcomes any comments.


Posted by: Jason Cawley

That is a really cool program. I played with some "star war" like rules, but not on a square grid. I found a few persistent groups in the 2 sizes of squares tiling (S5 in his files), with a 3 color rule with 2 -> birth, {3,4,5} -> survive, when usually it dies out. Add extra surival at 6 and it lives forever from typical initials (you get an array of large squares that fix at value 0 and lots of lively stuff between them, which eventually settles down to some "ko"-like blinkers). I haven't tried all the exotic tilings but it looks like a quite capable set up.


Posted by: Craig Piers

Thanks for finding the program. It's great fun! I came across something strange, though, while using the star wars rule on simple square array and wondered whether someone could explain it to me. I can't tell if it is a programming bug or whether it's just that I don't fully understand the effect of the "states" parameter.

To see it, load square tiles ("S1.til") and load "star wars.rul". In the rules editor, change "states" to "2". To start, set one cell in the array to 1 and its 4 corner neighboring cells to 1 ( :-: ). Take five steps, and remember that pattern. Now run the whole thing again with one change--change the "states" in the rules to "3". The strange thing is that the pattern of 1s is different at step 5 (loss of two 1s), a difference that results in the pattern dying out in a few more steps.

Any thoughts? What am I missing? Thanks.


Posted by: Craig Piers

In a private communication, Brian Prentice was generous enough to answer my question about his program. I thought I would post his answer for those of you who--like me--read these threads through and like a sense of closure, as well as for those of you who are just getting up to speed with CAs and shared my confusion.

My confusion centered on the states function, or "generations" parameter. Very simply, when a cell is in any state other than 0 or 1 it simply occupies space in the neighborhood and is not considered when deciding the survival or birth of its respective neighbors. As such, it decreases the size of neighborhood. For instance, if using a Moore neighborhood (8 neighbors) and one of the cells was in a state other than 0 or 1, there would be only 7 neighbors to consider when determining the survival or birth of the cell.

It should be understood, as well, that a cell assumes a state other than 0 or 1 only after it is a 1 and its neighborhood does not provide the conditions for its survival to the next step. This function basically allows you to watch previously alive cells slowly burn out. With each time step, the once alive cell changes state once. When it reaches the maximum state count, the cell returns to state 0 and is again included in the neighbor count as a 0.

Thanks, Brian. It’s a great program.




Forum Sponsored by Wolfram Research

© 2004-2009 Wolfram Research, Inc. | Powered by vBulletin 2.3.0 © 2000-2002 Jelsoft Enterprises, Ltd. | Disclaimer
vB Easy Archive Final - Created by Xenon and modified/released by SkuZZy from the Job Openings