Random points in U and V direction on surface!

Replies to This Discussion

Permalink Reply by Chris Tietjen on December 18, 2011 at 8:09am

Maybe like this?

Permalink Reply by Artyom Maxim on December 18, 2011 at 8:28am

Thanks for the answer, but no, as you can see, the setup creates the "rows" of points which is not what I need.

I need all the points to be at random positions on the object, but also I want to control the amount of them in each direction, U and V.

 

Permalink Reply by Artyom Maxim on December 18, 2011 at 8:30am

I've just thought, maybe I should create a desired distribution of points on a plane and then map them somehow to the surface?***

Permalink Reply by Hannes Löschke on December 18, 2011 at 1:26pm

see here

[edit]

having some kind of "grid" is never random.

try using a true grid of points and then moving the uv's by a random amount. that my be what you want to create.

Permalink Reply by Hannes Löschke on December 18, 2011 at 1:49pm

Attachments:

• RandomGrid.gh, 6 KB
• RandomGrid.png, 236 KB

Permalink Reply by Chris Tietjen on December 18, 2011 at 5:45pm

How about this?

Permalink Reply by Artyom Maxim on December 19, 2011 at 5:19am

Thanks for help guys,

Hannes, I've tried your def, but couldn't get it working right, however, I think it's close to what i'm looking for.

I'll go a bit deeper with my question- I'm trying to make a 2d voronoi division on a 3d surface, and as you can see on the second image, the cells are squeezed in Z direction, so I want to make more points in XY direction, to make the cells more uniform.

 

 I've added a cullpattern to get rid of those points on the top and bottom edge of the shape on the first image.

The editing of the two first sliders don't affect the cells proportions very much...

Ideas?

Thanks a lot. 

Permalink Reply by Hannes Löschke on December 19, 2011 at 5:34am

The EvaluateSrf will give you valid output even though the UV Koordinates are well outside your surface.

Assuming you wanted a controllable number of points in U and V direction, I pulled those Points outside the surface back to the edge (SrfCP).

The whole definition takes an evenly distributed grid of points and randomly displaces them. The domain component defines the distance of how far they will be moved form their point of origin. This should normally be a max of your individual grid size. In your case, this range seem to be way out of tune, as lots of points end up outside of your surface.

Try starting of with a range of 0 and incrementing it until you are satisfied wit the randomness of your grid.

Permalink Reply by Artyom Maxim on December 19, 2011 at 5:47am

 Could you please show me how to do that, because seems that I just dont get something in this def(

Attachments:

• random_grid_re.gh, 11 KB

Permalink Reply by Hannes Löschke on December 19, 2011 at 9:43am

I was wrong on the UV dimension. They range from 0 to 1 so the diplacement neets to be way smaller than 1.

unrolling the surface to generate 2D voronoi kinda makes the whole 3D random grid obsolete. so is took the freedom do implement it with the new 3D voronoi

Attachments:

• random_grid_re.gh, 10 KB

Permalink Reply by Artyom Maxim on December 19, 2011 at 10:45am

Thanks a lot, that's exactly how I would expect it to work! 

So you're saying there's no any way to feed those points to voronoi correctly?  

Permalink Reply by Hannes Löschke on December 19, 2011 at 2:08pm

there is... you'd have to create a 2d field of random points, because that's what 2D voronoi is made for, and then map it onto your cylinder...