pyblitzdg is an open-source physical modelling platform built on C++. The end-user writes their computational model in python 3+ with numpy, and leaves the sophisticated (often, boilerplate) details to the behind-the-scenes C++ implementations provided by pyblitzdg.
The pyblitzdg platform began as a pet project by Derek Steinmoeller, and it later evolved into one of the Waterloo Quantitative Consulting Group's core free open-source software initiatives.
The back-end of pyblitzdg focuses on robust and efficient C++ implementation details for Galerkin finite element methods with support for Discontinuous and Continuous Galerkin approaches in the python front-end.
From a historical point-of-view, most scientific physical modelling programs were written by researchers using low-level languages such as C or Fortran, for speed and memory efficiency. Since most numerical libraries only implement very low-level details not specific to finite element methods, this left a gap for most physical model developers to fill in on their own. As a result, it could take one or more years to complete model code development and establish, via test-cases, that the output was correct.
pyblitzdgalleviates these headaches by providing finite element implementations that are agnostic to their end use-cases in a model. These implementations are backed by unit-tests asserting the correctness of each independent piece of the bigger code puzzle. With a high percentage of code-path coverage and validating numerical assertions, pyblitzdg offers a base-line of openness and quality that you can count on for your model development work.
Pyblitzdg official docs by Waterloo Quantitative Consulting Group (WQCG, a not-for-profit corporation) is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Based on a work at https://wqcg.gitbook.io/pyblitzdg-official-docs/.
Permissions beyond the scope of this license may be available at https://github.com/WQCG/pyblitzdg-docs.