question about voronoi

Replies to This Discussion

Permalink Reply by David Stasiuk on March 12, 2013 at 1:59am

The "voronoi groups" component will let you do this exactly.  Just set set points roughly diagonally opposite each other nearish to the corners of your bounding rectangle into the G1 input, and the rest into the G2 input.

Permalink Reply by anthony yim on March 12, 2013 at 2:24am

thanks for your help 

i wonder if the bounding shape can only be rectangle ?

thanks

Permalink Reply by David Stasiuk on March 12, 2013 at 3:02am

well yes, you need to calculate your voronoi using a rectangular bounding shape, but it's quite easy to trim any curves outside of any new bounds you'd like to use.

Permalink Reply by Ola Jaensson on March 13, 2013 at 7:34am

Or you could automatically generate the points that, when fed into voronoi groups G1 input, would generate your bounding shape, then feed your points into G2. Like this:

I don't know why the bounding box needs scaling.

Permalink Reply by Leonardo Ali on February 19, 2017 at 12:42pm

Hi Ola,

I tried to implement the script you shared above with my bounding curve. Do you know why the Voronoi component won't work?

If you could take a moment to look at my situation, that would be most appreciated

Attachments:

• Voronoi_ConvexBoundingCurve.3dm, 70 KB
• voronoi.gh, 15 KB

Permalink Reply by Ola Jaensson on February 20, 2017 at 1:38am

Hej Leonardo,

Yes, there are three things you need to change:

1. The bounding curve must be a *convex* polyline, as it is a voronoi cell itself.

2. Right click the last BBox component and change it to "Union Box". That box is the rectangular boundary of the top-level voronoi diagram.

3. There are a couple of errors/assumptions in my definition; the original boundary corner points need to be included in the top level BBox and the center of scaling of that BBox should be its own center, not 0,0,0.

You could probably get nice cobwebs if you start with the voronoi diagram you get by taking the corner points of yout shape and then treat it as a network of elastic threads in the Kangaroo plugin.

Permalink Reply by Leonardo Ali on February 20, 2017 at 6:23am

Thanks for your help, Ola! It was greatly appreciated!

What I am trying to achieve is in fact in the link video below.

I have decided to move away from the voronoi and instead work with separate curves to form a grid I desire more, and offset these to form minimal surfaces running it through Kangaroo.

Do you know why this happening?

Thanks for your help again.

Best regards

Leonardo

Attachments:

• Geometry.3dm, 81 KB
• Minimal Surface.gh, 28 KB

Permalink Reply by David Rutten on March 12, 2013 at 4:50am

For now yes. The constraints on the boundary are the same as the constraints of individual voronoi cells. It has to be planar, consist of only linear segments and be completely convex. Even co-linear segments are not allowed.

I could make it work on convex polylines but until I do you'll have to use the trim post-process approach.

--

David Rutten

david@mcneel.com

Poprad, Slovakia