Fog Creek Software
Discussion Board




Array

I need an examples really exist of a three dimensional array storage device.

Doha
Friday, July 30, 2004

Check out "The Tardis". It's multi-dimensional.

http://www.geocities.com/~mjbrant/drwho/images/postcard1.jpg

TheGeezer
Friday, July 30, 2004

How many bits is that?

Li-fan Chen
Friday, July 30, 2004

I thought the inside of the Tardis wasn't in this dimension though, and so doesn't technically exist.

Matthew Lock
Friday, July 30, 2004

L- space. Ook.

trollop
Friday, July 30, 2004

Three dimesional arrays?

Man, even thinking about that sort of stuff will blow your mind!!!

Mr.Fancypants
Friday, July 30, 2004

Humm, all storage devices are physically three dimensional :)  However, most are logically one-dimensional, so why don't you come up with a mapping? If you want a 3x3x3, your just going to get an array[27] anyway.  You might as well just start that way.

Steamrolla
Friday, July 30, 2004

People have difficulty visualising an array with more than three dimensions, apparently.

Think of it as a tree is often more natural.

i like i
Friday, July 30, 2004

>Think of it as a tree is often more natural.

How do you mean that?  I guess I don't see a good fit for a tree to an n-dimensional thing.  I can see a graph used for that purpose (you can make n-dimensional objects out of graphs anyway), but not necessarily a tree.

Kalani
Friday, July 30, 2004

Technically a tree is just a specialized graph.  But who wants to split hairs?

Elephant
Friday, July 30, 2004

I have a weird way of imagining mega-multi-dimensional arrays:

1-dimensional: line ---------------------

2-dim: graph

3-dim: cube


4-dim: line of cubes

5-dim: graph of cubes

6-dim: cube made out of cubes


7-dim: line of "cubes made out of cubes"

...and so on.

not convinced
Friday, July 30, 2004

I saw a picture of that in a textbook years ago and have been plagued with the same image...  Seems to work though.

_
Friday, July 30, 2004

"Cube made out of Cubes.." -- Isn't that a 'tessaract'?

(..And he built a crooked house.  Heinlein.)
(Also, A Wrinkle in Time.  Madeline L'Engle)

AllanL5
Friday, July 30, 2004

not convinced: nice. I do the same thing, and almost started to try to explain it, but my explanation would've been about 10 times as verbose.

Visualizing n-dim arrays is sure a heck of a lot easier than visualizing n-dim spaces. I remember having tried to take an arbitrarily-shaped (but simple) object and imagine rotating it through a fourth dimension -- had to stop before my brain exploded and made a big mess all over the room. Has anyone managed to think of a reasonable way to visualize n-dim objects that are not broken into finite, immobile chunks the way array slots are?

John C.
Friday, July 30, 2004

Doha -
As you've probably figured out, this is not the place to be asking that sort of question.  Try Google.

John C. -
I get to the "cube of cubes" stage and can image the inner cubes rotating around each other.  But beyond that, I totally lose it.

Guess I'm just not cut out to be a Time Lord.
;-)

example
Friday, July 30, 2004

www.timecube.com

Jimmy Jo-jo
Friday, July 30, 2004

Elephant, but that's the point -- it's a more restrictive form of graph than is most obvious for the n-dimensional 'visualization'.  By definition a tree can't contain cycles but, by definition, an n-dimensional space does have cycles.

1D is a graph like ('o' == node, '--' == edge):

... --o--o--o-- ...

So in 1D, each node has degree 2.  In 2D, each node has degree 4 (or 8 if you're into 2D cellular automata), and so on.  You can use this method to construct the elements of any n-dimensional space (but the space has to be discrete, as somebody else mentioned).

Kalani
Saturday, July 31, 2004

I don't visualize arrays, I just see the dimensions as attributes: [city#][street#][street address][apartment wing#][floor#][room#] etc. There's six dimensions right there.

Bob
Saturday, July 31, 2004

About higher than 3 dimensions - a mathematician would probably tell you something along these lines:

'Visualising' higher-dimensional spaces in the conventional sense is hard/impossible for some - but if you think about them enough, about generalising various 2 or 3-dimensional problems to n dimensions, you end up looking past that and finding other ways of 'seeing' these spaces in a more abstract but equally valid way in your mind.

In short, if you look beyond the need to visualise things and really get your head round the idea of n-dimensional contructs from a formal/abstract point of view, in time you'll find yourself visualising them in a sense, just not the way you would have expected.

Visualising a 3-dim'l array should be a piece of piss though, just think about lots of little cubes arranged in a 3-dimensional grid.

Matt
Saturday, July 31, 2004

*  Recent Topics

*  Fog Creek Home