The GUDHI library is a generic open source C++ library, with a Python interface, for Topological Data Analysis (TDA) and Higher Dimensional Geometry Understanding.
I tried to make the code more robust, in particular in case the first d+1 points are affinely dependent (there is some discussion about different strategies at the end of the header).
I only did a dD version. While using CGAL's Triangulation_2/Triangulation_3 is indeed faster (because of the hierarchy and structural filtering, which both speed up point location, an important part since we cannot sort the points as in the usual Delaunay), the difference is not that huge on the data I tried, especially with CGAL-6 (I pushed a few simple optimizations there), and that part is cheap anyway, in 3d we typically spend more time inserting in the Simplex_tree than doing CGAL stuff...
Unlike most Gudhi files, I added a GPL license (+ exception for GeometryFactory customers) on this one
since it is unusable without CGAL, we don't lose anything by doing that
it contains code copied from CGAL and tweaked for this application
although I did not copy code from Michael Kerber's repository, I did look at it, and his code is GPL (he did not protest against the exception)
The details differ a bit from https://bitbucket.org/mkerber/function_delaunay/src/master/ , but it builds the same thing in the end.
Unlike most Gudhi files, I added a GPL license (+ exception for GeometryFactory customers) on this one
Documentation available at https://output.circle-artifacts.com/output/job/78531865-e05c-4d25-ba74-46ccbbd3b646/artifacts/0/doxygen/index.html