Hello!
I have:
- a 5 KB .txt file where each of the 85,000 lines represents a single 3D point given as x, y, z coordinates
- a 9 GB binary .npy file which, when loaded into numpy, is an 85,000 x 14,000 matrix of integers in the range [0, 255] where the element stored at (i, j) represents the color that point ‘i’ should render as at time slice ‘j’
- the color is grayscale i.e. the element at (i, j) is to be duplicated as each of R, G, and B when rendered in Rhino
I want:
- pretty low latency when displaying the point cloud at time slice j
- a slider that controls which of the 14,000 time slices is currently being rendered
What are your thoughts on how to do this?
Rhino becomes unresponsive when trying to load the 9 GB .npy file with CPython (my computer has 32 GB of RAM). Haven’t had much success with reading and parsing CSVs this large with Grasshopper either.
Possible to make use of Rhino’s point cloud optimizations somehow? I have a fancy GPU fwiw.
Thanks!!