New avatar

Post by **Tank Program** » Thu Aug 14, 2008 5:48 pm

"It is but it isn't."

*sigh*

Is it derived then from some sort of numerical algorithm? A random input from which the pattern is created. That would count as semi-random I think.

Jonathan · Post by **Jonathan** » Thu Aug 14, 2008 7:24 pm

The input is constrained. And if you provide your own input, it would look similar most of the time. Also, output is a meaningful interpretation of the input, and is very clearly related to it once you have a clue. It is even helpful in understanding the input.

Post by **Tank Program** » Thu Aug 14, 2008 7:44 pm

A case where the computer is more imaginative than people? I still think it's some sort of sequence or something.

Jonathan · Post by **Jonathan** » Thu Aug 14, 2008 8:20 pm

Tank Program wrote:A case where the computer is more imaginative than people?

More of a use of its raw power to help identify important parts of the input.

Tank Program wrote:I still think it's some sort of sequence or something.

I'd prefer to call it a set. You could shuffle it or even process items in parallel (provided you use proper locking and duplicate processing state for each thread, or duplicate everything and recombine trivially later).

Post by **Tank Program** » Thu Aug 14, 2008 9:53 pm

So if you had a range of say 1 to 10, you'd create some input like {3, 5, 8, 3, 9, 2, 6, 3, 2, 1} and that would give you your diagram. Except that somehow those numbers are linked, and while semi-meaningful in that state, the diagram gives them a better representation.

Jonathan · Post by **Jonathan** » Thu Aug 14, 2008 10:11 pm

Erm, no.

It's not arbitrary. And when I said the output is meaningful and helpful, I meant it actually guides humans through the input. Not just some weird graph that 'helps' because it's vaguely related.

Post by **epsy** » Thu Aug 14, 2008 11:12 pm

so it basically help us to see that these are 'near' of each other?
you previously said that it was some kind of map and a kind of set(hmm, two opposed types in STL

), are the 'coordinates' three-dimensional ?

Jonathan · Post by **Jonathan** » Fri Aug 15, 2008 1:23 am

It's a map as in map of a city. You feed my program the (constrained) city and you get the map.

It's all 2D.

Phytotron · Post by **Phytotron** » Fri Aug 15, 2008 7:22 am

Alright, so I must not recognise it after all. Too computer geek for an artist/nerd.

Your next avatar endeavor should be to create a representative visual manifestation of your geekified peacock feathers.

fantomzed · Post by **fantomzed** » Fri Aug 15, 2008 2:27 pm

Er... string????

Post by **Tank Program** » Fri Aug 15, 2008 6:19 pm

Is it a map as in some sort of array[][] input then or just 1d input for the 2d output there. Is there anything to do with intensities? Higher numbers at position [x][y] gives more white?

Jonathan · Post by **Jonathan** » Sat Aug 16, 2008 1:29 am

Input is also 2D. Input/output positions correspond directly to each other. Higher intensity means more of something the algorithm measures, but input has no sense of brightness.

Phytotron: great idea.

Post by **Tank Program** » Sat Aug 16, 2008 8:23 am

So it's a measure of something algorithmic, that's not related to light. Does its 2D coordinates have a spacial interpretation then?

Jonathan · Post by **Jonathan** » Sat Aug 16, 2008 10:37 am

Yep. The whole thing is very spatial.

Post by **Tank Program** » Sat Aug 16, 2008 12:27 pm

Outer-spacial? Visual-spacial?

And there's still that algorithm somewhere too right?