This is the repository for my thesis. If you'd just like to read the paper, feel free. Virtually all UC Berkeley dissertations are made available online via ProQuest; mine can be found in the UCB Library Catalog (PDF).
This repository is laid out in a manner described in Good Enough Practices in Scientific Computing.
The content itself has been broken into a few standalone papers and uploaded to the arXiv and / or submitted to journals:
K-Compensated de Casteljau paper (GitHub repo, published in AMC on April 5, 2019)- 2-Norm Condition Number for Bézier Curve Intersection (GitHub repo, published in CAGD on November 8, 2019)
- A Curious Case of Curbed Condition (GitHub repo)
- High-order Solution Transfer between Curved Triangular Meshes (GitHub repo, submitted to CAMCoS in October 2018)
In addition, this repository has slides for:
- Thesis talk
The code used to build the manuscript, generate images and verify
computations is written in Python. To run the code, Python 3.6
should be installed, along with nox-automation:
python -m pip install --upgrade nox-automation
Once installed, the various build jobs can be listed. For example:
$ nox --list-sessions
Available sessions:
* build_tex
* make_images
* update_requirements
To run nox -s build_tex (i.e. to build the PDFs), pdflatex,
xelatex and bibtex are required. In addition the metropolis
Beamer theme should be installed, as well as the Fira font family.