Archinect
anchor

Algorithms

popeye

I realize that an algorithm is defined as a computational procedure for solving a problem in a finite number of steps, but what does that really mean?
Algorithms are credited with aiding anything from animation software, to solving a rubricks cube, to creating voice recognition software.

Can someone explain algorithms to me like I am a 2 year old, there is just something I am missing?

Thanks, Denzel (a la Philadelphia)

 
May 29, 05 11:57 am
Cloutier

ummmmm Algorithms can be used to do pretty much anything (including architecture)

algorithms are really just instructions on how to solve a certain problem: a stupid example: an algorithm for a cat:

1. am i hungry?
no then go to step 3
yes then:

2. is there food in my bowl?
yes then eat, go back to step 1.
no then go annoy the humans, go back to step 2.

3. am i thursty?
no then go to step 6.
yes then:

4. is there water in my bowl?
yes then drink, go back to step 1.
no then:

5. is the bathroom door open?
yes then drink, go back to step 1.
no then annoy the humans, go back to step 4.

6. am i dirty?
yes then lick myself, go back to step 1.
no then:

7. am i tired?
yes then sleep, go back to step 1.
no then sleep anyways, go back to step 1.

this way, an endless cycle is created, by following this method the cat should never be hungry, thursty, dirty, or tired.

May 29, 05 9:52 pm  · 
 · 
donutboy

Algorithms are a set of instructions, based on discrete or logical mathematics. Condinationals (if, and, or, not) as well as loops (for, recursion) are the basic tools for which problems are solved (there are many many more.) The languages of computer science algorithms are quite varied, but the most common are object-orientated languagues such as Java, C++. Fuctional programming languages like LISP have only survived in the theoritcal world of comp sci (though I think CAD was written in LISP.)

May 30, 05 11:01 pm  · 
 · 
yalo

Wikipedia might be of some help:
http://en.wikipedia.org/wiki/Algorithm

If you are interested in some simple examples of algorithms applied to architectural design look here: http://designtooling.mit.edu
and here: http://www.shapegrammar.org/

May 31, 05 12:36 am  · 
 · 
bigness

wow, one of those questions i never had the courage to ask...

does anyone know of a book on algorithms for lay people,
on the level of Watt's "6 degrees" for the science of networks, if you know what i mean?

May 31, 05 7:40 am  · 
 · 
yalo

For a visual introduction to programming you might try John Maeda's book, Design by Numbers.

Bill Mitchell's book, The Logic of Architecture, provides a good conceptual introduction to the use of algorithms in architectural design.

Perhpas most helpful if you want to start writing your own code...here is a web-based reference on scripting by Stelios Dritsas: http://designtooling.mit.edu/components/sketching/scripting.html

May 31, 05 11:26 am  · 
 · 
badass japanese cookie

a cool band for which the former vice president played the synths

May 31, 05 11:10 pm  · 
 · 
bdonn

Interesting topic. I studied Comp Sci for 2 years before almost dieing.

However now that I've been studying architecture for a few years, I find a lot of the stuff I learned before really fascinating.

I recently heard about a design team that came up with an algorithm to design a table. (The legs were composed of diagonal beams etc...)

Consequently, every table was different.

May 31, 05 11:27 pm  · 
 · 

yeah it is an interesting approach, though algorithms harnessed to aesthetics always seem to me a bit of a fake approach becasue there is usually an arbitrary choice made by the designers to stop it when the forms are lookin particularly fine and the result is seldom as automatic as is often implied.

not that algorithms are bad mind, just that i think you can get more than aesthetics out of them. anyway, you can see the funky furniture guys here (http://www.kramweisshaar.com/frontend/index.htm ) if you are interested.

about the idea of algorithms in general i always recommended Darwin's Dangerous Idea, by daniel Dennett. There are books on emergence and so on out there as well but most are rather pop-ish and tend to be overly optimistic and absurdly dogmatic.

Jun 1, 05 8:01 am  · 
 · 
bigness

i read about the danish suo that was producing algorithm based tables, but it sounded a bit contrived, since they would let the computer run loose and generate 40 tables out of which they'd pick the 1 or 2 they liked. it's editing, is not designing...

it was in domus, march or april issue, in case anyone wants to look it up.

Jun 1, 05 8:11 am  · 
 · 
bigness

i read about the danish suo that was producing algorithm based tables, but it sounded a bit contrived, since they would let the computer run loose and generate 40 tables out of which they'd pick the 1 or 2 they liked. it's editing, is not designing...

it was in domus, march or april issue, in case anyone wants to look it up.

Jun 1, 05 8:11 am  · 
 · 
A Center for Ants?

shouldn't Per chip in about now?

Jun 1, 05 7:32 pm  · 
 · 

dam you romans!

Jun 1, 05 9:16 pm  · 
 · 

actually this guy does some nice work with math, and beats the shit out of m. per:

http://www.materialsystems.org/

Jun 1, 05 9:19 pm  · 
 · 
imagebytekid

Algorithms are used to describe.
They are mathematical 'metaphors' used to explain certain things that may be abstract or literal or both.
They can describe a shape or form such as a cube, a circle or a even complex form such as human face or architechtual facade.
In architecture an example would be:
http://www.labarchitecture.com/
see the fed square fractal facade...which looks fantastic espeically over the atruim entrance.
Here one is able to see the Algorithm in form and function:)


Jun 1, 05 10:49 pm  · 
 · 
Luis Fraguada

yeah, that is some pretty interesting stuff. I like how the site is layed out. The manifold project is well documented and shows a clear process from idea to prototype and beyond.

Jun 1, 05 11:02 pm  · 
 · 

labarchitecture is a pretty superficialexample though, sort of the version that charles jencks likes to talk about.

algorithms have the potential for a lot more depth than that, literally.

Jun 1, 05 11:08 pm  · 
 · 
Luis Fraguada

I was saying I liked materialsystems.org . . . been looking at that site, plus all of the people he links to . . . interesting work being done with these . . . EmTech and DRL at AA have some seemingly congruent studies in applicable algorhythmic systems.

Jun 1, 05 11:28 pm  · 
 · 
imagebytekid

hmm true...nice site....
although at least Lab built it, and it seems to have more form than the Babel concept on materialsystems.org.
That's what gets a rise out of Jencks...


Jun 2, 05 12:47 am  · 
 · 

well if we are gonna get into jencks, my big peeve with him is that he is claiming that there is a universal truth that complexity theory can somehow grasp. This is of course just the same old modernist rhetoric dressed up with a frilly dress. but it is worse becasue it makes the calim that decoration actually determines appropriate design. He never talks about program or inhabitation and seems to think the right backdrop is enough.

i like the stuf coming out of AA's lab because it drops this kind of argument altogether, making no claims for humanity and just getting on with examining the possibilities. No a prioris here thank you.

ito is now doing a project is scotland with cecil balmond that takes things in a halfway direction betwen the two extremes of graphics and formmaking. column placement and facade treatment are both controlled by algorithms, the first linked to function the other more or les decorative. it is sort of ugly but should be interesting to see when done. i think it is in CD phase now.

Jun 2, 05 1:09 am  · 
 · 
imagebytekid

Hmm yes.
content and marketing methinks.
There is plenty of good content...however,
the marketing is better when there's not:)

Jun 2, 05 1:50 am  · 
 · 
Mason White

(unfortunately, i think that the Ito/Balmond Selfridges in Trongate, Scotland was ditched when Selfriges was bought out by the Canadian company... at least that was the news last january.)

Jun 3, 05 12:10 am  · 
 · 

that figures. it was in the latest A+U but i only scanned the text. looked like it was just about ready for bidding... ah well.

the text i did read included an amusing exchange between ito and balmond that was priceless; cecil telling ito to stop being so anal and ito wondering aloud if cecil's problem was part of the western psychosis and therefore not his fault. i think balmond won the argument in the end. pity about the building, if true.

Jun 3, 05 2:21 am  · 
 · 

Block this user


Are you sure you want to block this user and hide all related comments throughout the site?

Archinect


This is your first comment on Archinect. Your comment will be visible once approved.

  • ×Search in: