WhatsThePoint.jl

A Julia package for generating and manipulating point clouds for meshless PDE methods — RBF-FD, generalized finite differences, SPH, and related techniques. Part of the JuliaMeshless organization.

Full documentation

Note

WhatsThePoint.jl is under active development. The API may change before v1.0.

Features

• Surface import from STL and other mesh formats via GeoIO.jl
• Volume discretization with multiple algorithms:
  • SlakKosec and VanDerSandeFornberg (3D)
  • FornbergFlyer (2D)
  • Octree — octree-guided adaptive point generation (3D)
• Octree-accelerated spatial queries via TriangleOctree for fast point-in-volume testing
• Normal computation and orientation using PCA with MST+DFS consistent orientation (Hoppe 1992)
• Node repulsion for optimizing point distributions (Miotti 2023)
• Point connectivity with k-nearest neighbor and radius-based topology
• Full unit support through Unitful.jl
• Visualization with Makie.jl

Installation

] add https://github.com/JuliaMeshless/WhatsThePoint.jl

Quick Example

using WhatsThePoint, Unitful

# Import a surface mesh
boundary = PointBoundary("model.stl")

# Split surfaces by normal angle
split_surface!(boundary, 75°)

# Generate volume points
spacing = ConstantSpacing(1u"mm")
cloud = discretize(boundary, spacing; alg=VanDerSandeFornberg(), max_points=100_000)

# Optimize point distribution
cloud = repel(cloud, spacing; β=0.2, max_iters=1000)

# Add point connectivity
cloud = set_topology(cloud, KNNTopology, 21)

# Visualize
using GLMakie
visualize(cloud; markersize=0.15)

See the documentation for the full guide.