Software
The package is available on CRAN.
A toolbox for systemic risk based on liabilities matrices. Contains
a Gibbs sampler for liabilities matrices where only row and
column sums of the liabilities matrix are observed. It is
based on the following paper:
Gandy, Axel and Veraart,
Luitgard A. M., A Bayesian Methodology for Systemic Risk
Assessment in Financial Networks (March 19, 2015). To appear in Management Science. Available
at SSRN
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2580869
This package is available on CRAN. It contains a short example.
An implementation of the chopthin-algorithm for resampling of particles described in
G&Lau (2015). The chopthin algorithm for resampling. arXiv:1502.07532
The implementation is C++ based
(chopthin_C++.zip contains the main code and an example). Also available are
an R-package
(chopthin on CRAN) a python package and a MATLAB MEX file
(chopthin_MEX.zip).
These functions implement the method described in the Appendix of Gandy, Kvaloy, Bottle & Zhou (2010), Biometrika 97 (2): 375-388. The implementation is written in R.
It is available in a git repository on Bitbucket
This package provides efficient MCMC algorithms for uniform sampling of graphs conditional on vertex level data. It is
based on the following paper:
J. Scott and A. Gandy. State-dependent kernel selection for conditional sampling of graphs. Journal of Computational and Graphical Statistics, 2019. Accepted for publication. arXiv:1809.06758 [stat.ME].
This package is available on GitHub.
This package provides unit tests for MCMC and Monte Carlo methods. It is
based on the following paper:
A. Gandy and J. Scott. Unit Testing for MCMC and other Mone Carlo Meethods, 2020. arXiv:2001.06465 [stat.ME].
This package is available on CRAN and on bitbucket.
Please send comments to the e-mail address given on the main page.