algorithmic modeling for Rhino
Hi Panagiotis and Sawako,
Thank you so much for sharing your amazing plug-in! It is so exciting and I love the graphics too :) Sorry I already have two very basic questions:
-Is there a way to run millipede on both Rhino 4 and 5?
So far I have it only on 5 as i am using the 64 bit version, do I need to always switch between the two gha files in my component folder?
-I am trying to understand how to use the Isosurface component but struggle to find what a "3D grid resolution" is. Is it the x,y,z parameters of a f(x,y,z) function?
If i have a random list of points in space (which did not come from a function), how do I extract their resolution?
At the moment the only way to run both versions is to replace the files when switching from 4.0 to 5.0. The problem lies with sawapansolvers.dll which links against different versions of intel math kernel libraries. I am working on a more elegant solution but at the moment it is final exams months and will not be able to do much.
Iso surfaces are calculated using the marching cubes algorithm and as the name suggests the algorithm operates on a three dimensional grid [not necessarily cubic]. You don't need to input the actual points just the scalar values that correspond to the value of the F[x,y,z] function. But htese values must be defined on a grid of resolution resX x resY x resZ. If you see the example that computes the isosurface it does exactly that. It evaluates a function at points on a grid and outputs just the values of the funciton not the points.
If you have an unstructured set of points then depending on what you need to do there are different options. If you just want to wrap a surface around the points [blob style] then just create a 3d grid enclosing the points and calculate the sum of the gaussian funcitons of the distances to all point at each grid point and this is your F[x,y,z] function. If your points represent random sampling of some F[x,y,z] funciton in space then you cannot use this plug in. You probably would like to do something like tetrahedralize the space between the points [3d delaunay] and then apply the marching tetrahedra algorithm to the result to get the iso surfaces [marching tetrahedra is even simpler than marching cubes]
I want to create some 'blob style' forms as shown in the attached image (this was created in 3DS max), however I would like to create this in Rhino/GH so that I can eventually form a gridshell structure. In your reply to Arthur above you stated:
"If you just want to wrap a surface around the points [blob style] then just create a 3d grid enclosing the points and calculate the sum of the gaussian funcitons of the distances to all point at each grid point and this is your F[x,y,z] function"
Based on the image attached which method would be best to use? The forms are relatively arbitrary.
Any help you can give me would be much appreciated.
Hi Panagiotis and Scott,
Thanks to your help and the IsoAlgae definition of Marios Tsiliakos from Digital [sub]stance I managed to create an Isosurface from a list of points. As Panagiotis suggested, this is done through wrapping a surface around the points using the "Geometry Wrapper" component. Please see attached files and printscreen below (for rhino 5).
Best of luck for your exams Panagiotis!
Hi I am really trying to do something similar to here. i want to use isosurface for flocking, (so use this to trace the trail) however i really dont know where to start. i read about millipede and seen a lot of example but i still don't understand...any suggestions lease help.
could I use geometry wrapper with variable radius of the wrapping? that means either use a graph or an attractor to manipulate the size of the wrapping ?
I can t get millipede to run on the latest version of rhino 5... Does anyone know how to fix it....downloaded the 64x libaries.
just uploaded a new zip file that should work on both 32 and 64 bit [rhino 4 and 5]. At least it works on my machine. You need the latest Rhino 5.0 and the latest grasshopper