WhatsThePoint.jl

Appearance

WhatsThePoint.jlPoint clouds for meshless PDE methods

Generate, optimize, and connect point clouds — from CAD surface to solver-ready in a few lines of Julia.

Get Started
Guide
API Reference
View on GitHub
Discretized Stanford bunny point cloud with cutaway showing interior points
📥

Import & Sample Surfaces

Load STL, OBJ, or any GeoIO.jl format with explicit units, then Poisson-disk sample the surface at exactly the spacing you want.

🫧

Fill the Volume

Four discretization algorithms generate well-spaced interior points, including spacing-driven adaptive orthtree fill with graded Poisson-disk fronts.

🧲

Optimize the Distribution

Node repulsion relaxes point clouds toward blue-noise quality, with boundary-aware projection and quality-based stopping.

🕸️

Build Stencil Connectivity

k-nearest neighbor and radius-based topologies give every point the neighborhood your meshless solver needs.

📐

Units & Quality Metrics

Full Unitful.jl integration and built-in separation, fill, and spacing-fidelity metrics keep discretizations honest.

⚡

Fast by Default

Threaded operations via OhMyThreads.jl, cache-friendly StructArray storage, and orthtree-accelerated O(1) point-in-volume queries.

Quick Example ​

Load a surface mesh and inspect the boundary structure:

julia
using WhatsThePoint
using Unitful: m
mesh = import_mesh(joinpath(@__DIR__, "assets/bunny.stl"), m)
boundary = PointBoundary(mesh)
PointBoundary{Meshes.𝔼{3}, CoordRefSystems.Cartesian3D{CoordRefSystems.NoDatum, Unitful.Quantity{Float64, 𝐋, Unitful.FreeUnits{(m,), 𝐋, nothing}}}}
├─69664 points
└─Surfaces
  └─surface1
julia
using GLMakie
visualize(boundary; markersize=0.15)

Generate volume points with discretize:

julia
spacing = ConstantSpacing(1m)
cloud = discretize(boundary, spacing; alg=Orthtree(mesh))
julia
visualize(cloud; markersize=0.15)

Installation ​

julia
using Pkg
Pkg.add("WhatsThePoint")