 |
Author |
Topic: cellular automata (Read 1344 times) |
|
steve rowe
Guest
|
cellular automata
« on: Aug 13th, 2002, 8:18am » |
|
a controversial new book by stephen wolfram called *A New Kind Of Science* explains wolfram's use of cellular automata to explain various scientific phenomena, but I also see possiblities to create original art using cellular automata. Wolfram's Mathematica software can be used to create one's own cellular automata, but I have not been able to try it - it's rather expensive (although student price is reasonable). Anyone interested in using algorithms to create art (as down with Proce55ing or Design By Numbers) might also find this worthwhile. I would love to hear from anyone who has read the book or used the software.
|
|
|
|
|
tomek
|
Re: cellular automata
« Reply #1 on: Aug 13th, 2002, 3:54pm » |
|
I just started reading the book. Mixed feelings so far. It is definetely interesting from the visual (graphic arts) point of view as it describes large number of computationally created patterns that simulate ones found in nature. The book got awful reviews from readers on Amazon, but I find that the ideas Wolfram writes about are interesting to ponder independently if his theories are correct or not.
On the topic of applying his ideas to computational design I find it would be probably very interesting, but also limited to creating complex pattern backgrounds, textures, textile designs and similar. Essentially instead of using random functions you could use his formulas to generate random looking textures but having significant (non random) control over their appearance.
As I said, I have not finished reading the book, so these are my initial observations, not a final review. There is a companion website to the book that I believe provides the programs for free download and samples of the book, so you might get plenty of information without or before spending any money: www.wolframscience.com
|
|
|
|
|
steve rowe
Guest
|
Re: cellular automata
« Reply #2 on: Aug 13th, 2002, 11:35pm » |
|
tomek, thanks for the reply. you may be right about the limitations on using this stuff for graphics, but if wolfram's correct, than one should be able the do literally everything with cellualr automata - his basic claim is that, in fact, everything IS done with them. (I am somewhere in chapter 8, so don't have the full gist just yet.)
even given the apparent limitations of simple schemas, I would like to give them a try, but not at the expense of the whole Mathematica application. I checked out the web site, but find no downloadable apps. I guess i'll have to wait for Proce55ing 1.0, which looks like it will be worth the wait.
shrew
|
|
|
|
|
tomek
|
Re: cellular automata
« Reply #3 on: Aug 14th, 2002, 1:37am » |
|
OK, I was not clear enough. Here is more info:
1. From the mentioned previously website you can download a free MathReader. This application can open .nb (Mathematica notebook, I think) files. The 'downloadable programs' section of the site has a notebook I believe for each chapter of the book. I am not sure how interesting it is, but some other demos of .nb files on the same site have graphics embeded in them and (I think) you can run animations inside as well. That's all for free.
2. They already have A New Kind of Science Explorer that you can order for $65 that is something like a graphics engine with interactive input so you can run the book experiments and try your own modifications. I am not sure how useful that is but it is cheaper than Mathematica license...
3. Programming the experiments described in the book in Proce55ing is doable in current version and fairly easy (that obviously depends if you have any programming background or not). For example you can use setPixel(), getPixel() and pixels[index] array to implement them directly from the book descriptions. If I get around to play with it I will post an example, but I encourage you to try for yourself even if you're just learning how to program. And the best part is that this solution is free as well thanks to the folks of Proce55ing.net!
General note about the book: yes, the leading thought that EVERYTHING is computable using his way of thinking is an interesting concept, but at this point I am not convinced yet that he is right and even if he is then we need to understand 'his science' better to become fluent at using it in practical applications including computational design. For example if I want to add to my composition a circle I can do it immediately using 'traditional' programming and the known math formula while it might take me forever to try to find the right set of cellular automata rules... So in this sense it is (in my opinion) limited in applying to computational designs - it is difficult to predict what you will get when you run a program based on his ideas. So, it is a bit of a random process from the designer's point of view - you're not really in control of the end result. Unless, as I said you want to create texture like patterns based on the research that he already presents in the book and some of them are quite beautiful, I admit.
|
|
|
|
|
eviltyler
|
Re: cellular automata
« Reply #4 on: Aug 14th, 2002, 3:29pm » |
|
i have a new kind of science too. i'm pissed at myself for not being able to get past his writing style. he thinks a bit too highly of himself.
this review caught my eye, as it sums up the air of wolfram quite well, the review is at the bottom:
http://www.math.usf.edu/~eclark/ANKOS_humor.html
that being said, celluar automata is quite powerful. however, the idea of it being the base of everything is hard to swallow. but it's worth investigating and i think that proce55ing is a good enviro to start working on it.
i'm looking forward to seeing your progress/results.
|
|
|
|
|
shrew
Guest
|
Re: cellular automata
« Reply #5 on: Aug 15th, 2002, 11:38pm » |
|
tomek, thanks for the lead on the downloadadble software. of the 3 .nb's i downloadad, however, there appear to be no graphics, just notes and code - it makes you wonder why they make it so hard to read these files, as simple text files would have been easier then having to install the mathreader, and would have been just as effective. i guess it's for the loyal mathematica crowd. (btw, on my mac i had to open and save the .nb's with a text editor (bbedit) to get mathreader to recognize them. then it screwed up my mac pretty good, causing the finder to quit and restart repeatedly until i forced a restart - a trick i have rarely seen in my 14 years on that platform.)
as for the explorer software, it looks like a cheap way to see some automamta in action, but it grants a license only for personal use of the resulting graphics (imagine photoshop doing that). also, it's not available yet, but they will take your money now.
all very disappointing. i hope proce55ing can restore my faith - when is 1.0 expected?
|
|
|
|
|
ezdac
|
Re: cellular automata
« Reply #6 on: Aug 29th, 2002, 5:51pm » |
|
Hi i just discovered proc55ing, and i'm also highly interested in cellular automata, as a way to produce deterministic chaos (makes more sense than pseudo-random, which is hidden chaos). Thanks for taking my attention to Wolfram's book, i'd like to recommend you 2 softwares that demonstrates the "creativity" of such algorithms
http://www.cafun.de
http://hws3.hws.edu:9000/eck/index.html/
one is for PC, one is for mac, one is 1D, the other 2D, both java.
I'm trying to make waveforms from cellular automata and chaos, with max/MSP mostly, but it's not really adapted to recursive calculation, maybe proce55ing 1.0 will implement sound ?
good day !!
raphaël
|
|
|
|
|
|
