adaptmesh 0.1.2

Adaptive mesh generation and refinement

adaptmesh

Create triangular meshes by the adaptive process.

This is work-in-progress. I'm currently writing a short whitepaper which explains the logic behind this simple mesh generator. The basic idea is the use standard techniques from adaptive finite element methods, an arbitrary initial mesh, and a robust mesh smoothing algorithm. The next step is to make this more usable, i.e., add support for holes and tagging of boundaries. In future I hope to extend the generator to tetrahedral meshes. Stay tuned.

Installation

pip install adaptmesh

Dependencies

• numpy
• scipy
• matplotlib
• scikit-fem

Examples

Square with default settings

from adaptmesh import triangulate

m = triangulate([(0., 0.),
                 (1., 0.),
                 (1., 1.),
                 (0., 1.),])

Non-convex shape

from adaptmesh import triangulate

m = triangulate([(0.0, 0.0),
                 (1.1, 0.0),
                 (1.2, 0.5),
                 (0.7, 0.6),
                 (2.0, 1.0),
                 (1.0, 2.0),
                 (0.5, 1.5),], quality=0.95)  # default: 0.9

Licensing

The main source code of adaptmesh is distributed under the MIT License.

adaptmesh ships with customized versions of the following packages:

• tri v0.3.1.dev0 (ported to Python 3; MIT)
• optimesh v0.6.2 (trimmed down version with minor changes to the edge flipping; the last version with MIT)
• meshplex v0.12.3 (trimmed down version with minor changes, i.e. removal of unnecessary imports; the last version with MIT)

The licenses of the included packages can be found also in LICENSE.md and the respective subdirectories, i.e. ./adaptmesh/*/LICENSE. See LICENSE.md for more information.