Grasshopper

algorithmic modeling for Rhino

From what I read on the forum, the only way to move mesh vertices is to explode the mesh, transform the vertices that need to be moved, and then rebuild a mesh with the new vertices.

Isn't that like rebuilding the whole house when you just want to move a chair ?

This seems awkward...

Could someone help me with an example definition because I have no idea how to do rebuild the new mesh.

Thanks !

Views: 3848

Replies to This Discussion

Simply this:
Since rhino mesh is mainly just a list of points and a list of faces, this is not too inefficient (if there is no dozens of thousands of points). You're also accessing to normals and colors, adding some inefficiency. Grasshopper has no other way to solve this. In code, you would access only and specifically to the index of the list of points to modify it and replace it. In Grasshopper you need access to the whole list of points, modify some or all points, and then replace them all. It is more inefficient, but I think completely valid (in terms of runtime).

Hi Daniel,

Thanks for your example, but you are treating the easy case where you move the whole bunch.

If you want to move a subset of the vertices, it becomes pure inferno (well, for my tiny brain anyways).

In my case, I have an open mesh and I want to project the vertices which are on the naked edges onto a surface.

Here's my first attempt, but somehow, the "construct mesh" component is not liking what I am feeding him :(

Attachments:

The difficult in your case is to know the naked vertices. 

Attachments:

Got it : one of the points is not projecting because it is underneath the surface and I project "downwards".

If I use "Pull point" it works.

Phew... Having to use "Tree branch", "List length", "Series", "Cull index" and "Merge" is a testimony to the fact that moving a couple of mesh vertices is a tad cumbersome...

replace items does the job too.

best

alex

If you use the Point Deform component, you can operate on 'control points' directly. Since the order no longer matters (control-points are found by location, not index) it becomes a geometric rather than a topological problem.

Attachments:

Hi David !

I just discovered "Point deform".

Thanks !

By the way, I wonder why Components like "List idem" or "Cull index" don't have an option to keep the index of items constant, replacing unwanted items by a "Null".

There is loss of valuable information here...

This would make it easier downstream to merge the lists again, instead of having to graft and juggle with branches.

What do you think ?

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