Grasshopper

algorithmic modeling for Rhino

sets depending on overlapping (intersection) geometries

hey everyone!

looking for ideas on where to start to solve this:

I have several geometries, some of which overlap, i would like to sort the resulting regions in sets depending the number of overlapping geometries that formed them for each part. (ex: 1 intersection, 2 intersections...)

darker green means more intersections,

I want to evaluate intersection between all geometries but no with itself... can't even seem to be able to start

this is my gh. so far.

any ideas? is it possible?

the problem is that i need to sort the intersecting parts, make them new geometries, those would be sectors of my islands, and group them by type.

like in this image

i have an idea but it's not really working, use cross reference to make all the possible region intersections with all the geometries. and then evaluate their area center.

use that with member index to just draw one but sort them in sets/groups of type

I guess it's kind of a mess right? 

cheers,

Views: 1598

Attachments:

Replies to This Discussion

I have an idea, but it's somehow not working....

imagine I have 4 geometries, i need to do the region intersection of geometry 1 with itself and all the others, and so on... but gh seems to only do the intersection of all together, not one by one

Does this Work?

Ps You don't need to make them surfaces for the Region Commands to work they require curves, it takes your surface and just takes the edges curves.

Attachments:

thank you!! it does something super similar to what I need, I will rework and repost my results. Let's see if I can make the changes myself. 

hey!

yeah, no need to use surfaces, thanks.

i'll call those diamond shaped geometries islands.

your definition helps determine which islands have intersections and how many.

the problem is that i need to sort the intersecting parts, make them new geometries, those would be sectors of my islands, and group them by type.

like in this image

i have an idea but it's not really working, use cross reference to make all the possible region intersections with all the geometries. and then evaluate their area center.

use that with member index to just draw one but sort them in sets/groups of type

I guess it's kind of a mess right? 

cheers!

Hi.

I'm not sure this is what you want. Check attatched.

Attachments:

thank you Hyungsoo for your reply!

I will rework my definition with your help and repost my results tomorrow!

hey Hyungsoo! love it how you always use colours to identify sets.

see what I just posted above, your definition also helps identify groupss depending on how many intersections is the max that group has. but not redifine the whole geometry.

any ideas?

just used your colours for my example

cheers!

Here's a possible solution, it's slow because of the surface splitting but it works fine

Attachments:

andrew! you're the man!

it works like a charm, and using point groups and threshold to find overlapping geometries works so much better than member index that i was trying to use...

thanks a lot!! 

here's a screenshot of what it does, it might help someone else too

andrew, I did some adjustments to your solution to fit my case

I understand you did the collision test to save calculation time, but there might be some cases where geometries don't collide but areas overlap, like geometries inside other geometries.

I added a graft and flatten to the boundary edges to sort those in, and deleted the collision test.

there's only one problem not solved, that so far it doesn't affect my needs:

when there's two geometries that share the same area center, concentrical geometries...

Attachments:

RSS

About

Translate

Search

Photos

  • Add Photos
  • View All

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service