Region Intersection ruins Voronoi

Hey,

i'm trying to get a voronoi structure on an asymmetrical 4-Pt-Srf created in GH from a Rhino Curve. Everything works fine (even the voronoi) untill I get to the RInt part, which destroys the voronoi (adds cells without points, goes past the bounary etc.).

Been on it for hours and can't figure it out, does anyone know whats going wrong?

Thank you,

Randy R

Attachments:

Triangle.tif, 465 KB
Triangle.3dm, 152 KB
Triangle.gh, 19 KB

Replies to This Discussion

Permalink Reply by David Rutten on September 6, 2015 at 10:49am

Yeah something fishy is going on with the Region Intersection component.

I'm trying to find the bug now, in the meantime I've simplified your file a bit.

Attachments:

Triangle.gh, 8 KB

Permalink Reply by David Rutten on September 6, 2015 at 11:02am

No wait! Region Intersection operates on two collections of curves. If all the curves in one collection have no shared overlaps, then obviously the end result will also be empty.

Permalink Reply by David Rutten on September 6, 2015 at 11:08am

Attachments:

Triangle.gh, 10 KB

Permalink Reply by David Rutten on September 6, 2015 at 11:15am

And here's an approach that doesn't operate on all edges sequentially. It does however have some cells that are weirdly cut by the inner triangle:

Attachments:

Triangle2.gh, 9 KB

Permalink Reply by Randy Randelson on September 6, 2015 at 11:54am

First of all, thank you very much for the quick replies!

I tried a lot of options already to, thats why at the end the definition was so unnecessarily complex, cause I thought the voronoi or RInt would work better in smaller parts - but as soon as i set the number of points higher (even with your definition) cells appear that are not like voronoi cells

Is that still a problem with the logic or the program?

Attachments:

Triangle 2.tif, 395 KB

Permalink Reply by David Rutten on September 6, 2015 at 1:43pm

I think that's a problem with your tolerance settings. The voronoi component assumes that edges shorter than the Rhino document tolerance were never meant to be edges, so it collapses them. Furthermore the RInt component also uses the absolute tolerance settings to determine whether some resulting shape is 'too small to matter'. If you increase your tolerances (i.e. use smaller numbers) then press F5 to force GH to recompute I think it will work better.

Permalink Reply by David Rutten on September 6, 2015 at 1:43pm

A good rule of thumb is that your absolute tolerance setting should be roughly ten times smaller than the smallest detail in your geometry.

Permalink Reply by Randy Randelson on September 6, 2015 at 2:24pm

Great, works perfectly, thank you very much!

RSS