weird thing about the random component...

i wanted to start with a 3d grid and just move the points a bit around so that they aren't as regular.

Thing is, even though each point has a different associated seed value, the points turn out describing a diagonal in all views, as you can see in the pic, or in the 3d, if you download the definition.

The points do have random XYZ values when compared to their initial positions, but compared one to the other have very particular positions, as you can see. There seems to be a step size for the different random values.

Any idea WHY this happens, and how it can be avoided?

Thanks!

definition download: slightly%20randomized%20grid.gh

Replies to This Discussion

Permalink Reply by Danny Boyes on October 31, 2011 at 3:22pm

I would throw in some jitter components with different seed values to get a proper random list.

Permalink Reply by Andrei Raducanu on October 31, 2011 at 5:31pm

yeah, i thought of that, and it gets a decent result,, but it's not "proper" random, but a workaround.. :)

cheers anyway!

Permalink Reply by Danny Boyes on November 1, 2011 at 7:33am

How is it that a randomly ordered list of seed numbers isn't "proper". If its better surely that's what counts here.

Permalink Reply by David Rutten on November 1, 2011 at 5:03am

If it's diagonal it means the x and y deviation is the same. You should probably generate different random values for x and y offsets.

David Rutten

david@mcneel.com

Poprad, Slovakia

Permalink Reply by Andrei Raducanu on November 1, 2011 at 6:31am

every single coordinate used is based on a different seed value, and i checked and they are all different, yet they exhibit a step size..

(another interesting thing i found is that if you use the same seed value for different domains of equal lengths, the returned number has the same relative, position so to speak - for example, if i use seed 2 for 0 to 1 and 1 to 2, the returned numbers are 0.771094 respectively 1.771094, but again, in my example i used different seed values for each and every coordinate used..)

Permalink Reply by David Rutten on November 1, 2011 at 7:21am

Ok, fair enough. You're generating your seeds in a very predictable fashion though. This is a great way to expose the "pseudo" in the pseudo-random-engine.

Try the attached.

David Rutten

david@mcneel.com

Poprad, Slovakia

Attachments: