The following software packages are associated with my research projects. They are essentially free to be used, but the author(s) reserve copyright.
Stein Thinning
Post-processing of Markov chain Monte Carlo output using kernel Stein discrepancy as an optimality criterion.
Control Functionals for Monte Carlo Integration
A series of papers that begun with Mark Girolami and Nicolas Chopin in the Journal of the Royal Statistical Society, Series B, 2017.
Wendland Compact Support Radial Basis Functions
This short script uses symbolic integration to compute Wendland's polynomials, returned as a function handle that can be easily used.
Bayesian Probabilistic Numerical Methods for PDEs
These notebooks demonstrate how computations were performed for a series of papers on Bayesian probabilistic numerical methods. Joint work with Jon Cockayne, Tim Sullivan and Mark Girolami.
Bayesian Conjugate Gradient Method
This Python script executes the Bayesian Conjugate Gradient Method, a variation on the standard conjugate gradient method for solution of large linear systems that additionally provides uncertainty quantification. Joint work with Jon Cockayne and Mark Girolami.