Limiting max dimension of voronoi cells

Hey, first post. I hope there's a simple solution to this and it's just my inexperience with GH.

In generating a 2D voronoi bounded by a square, is there a way to control the max dimensions of the cells? I am going to be laser cutting and have a bed size of 24" x 48". (It's a 12' x 12' cube, each cell is fabricated individually)

There are a few solutions I would see as acceptable but I can't figure it out:

1. Be able to see the bounding box for each cell and I can manually control the number of points and the spacing to visually check each fits in my machine.

2. Be able to actually limit the cells to the max dimensions of the bounding box and have it be an "incomplete" voronoi diagram until I have enough points properly spaced.

3. Be able to have GH determine a minimum number of points and spacing so that the square is complete and the cells fit within a bounding box (my machine bed).

Any help here would be great. Let me know if any additional information can help, and if this question has been answered too many times.

Attachments:

issue-01.jpg, 32 KB

Replies to This Discussion

Permalink Reply by Ola Jaensson on March 26, 2013 at 12:25pm

Using Andrew Heumanns minimal bounding rects from http://www.grasshopper3d.com/forum/topics/simple-min-2d-bounding-bo..., you could analyze the voronoi cells and mark the ones that are too large.

Then you could either change the ranom seed for the population, increase the number of points or bake the points and adjust them manually. Or use the cell centers as points for a new voronoi diagram. Or something else.

I used Voronoi Groups because I've had bad luck with the ordinary 2d voronoi recently.

Permalink Reply by Patrick Beseda on March 26, 2013 at 3:05pm

This is great. Conceptually, whats the difference between the 2D voronoi and voronoi groups? Did you plug two inputs (the X and Y) into the sort function on the right?

Thanks so much. Now I'm working on exploding it and nesting it on sheets to cut. Ola, you're great. Thanks for taking the time today.

Permalink Reply by David Rutten on March 27, 2013 at 1:39am

2D Voronoi creates a single voronoi diagram in a rectangle boundary. Voronoi Groups is a recursive process. First it creates the top-level voronoi diagram similar to 2D Voronoi. Then it moves on to the next layer of points supplied by the next input. For each point it figures out in which cell of the first layer it belongs, then it creates a new diagram of only those points within that cell. Repeat for any additional inputs you may have defined.

It's basically voronoi diagrams within voronoi diagrams.

David Rutten

david@mcneel.com

Poprad, Slovakia

Permalink Reply by Ola Jaensson on March 27, 2013 at 2:56am

Is there a reason Voronoi Groups seems to deliver more stable results than 2D Voronoi? It doesn't seem as sensitive regarding the framing rect.

Permalink Reply by David Rutten on March 27, 2013 at 4:17am

Shouldn't, they use the same algorithm.

David Rutten

david@mcneel.com

Poprad, Slovakia

Permalink Reply by Ola Jaensson on March 27, 2013 at 7:25pm

This is what I often get:

Any idea why?

Not a big problem, I'll just continue using VorGroup.

Permalink Reply by Ola Jaensson on March 27, 2013 at 2:54am

I plugged the X and Y to the sort (and then dispatched them) as an easy way to make sure I compare the shortest side of the bounding rect with the shortest side of the cutting table. There may be more optimal ways, but that was the first method I came to think of.

RSS