RcppBASTA

Rcpp implementation of Fryzlewicz and Subba Rao's (2014) binary segmentation for transformed auto-regressive conditional heteroscedasticity (BASTA) algorithm to detect multiple-change-point in auto-regressive conditional heteroscedastic processes.

Installation

If you have the remotes package installed, then you can use the following to install RcppBASTA:

# Uncomment the next line if you don't have the remotes package installed.
# install.packages("remotes")
remotes::install_github("sjvrensburg/RcppBASTA")

Acknowledgements

Many thanks to Prof. Piotr Fryzlewicz who made their R implementation of BASTA freely available. You may view the original code at (http://stats.lse.ac.uk/fryzlewicz/basta/basta.html)[http://stats.lse.ac.uk/fryzlewicz/basta/basta.html].

PLEASE CITE the paper "Multiple-change-point detection for auto-regressive conditional heteroscedastic processes" by Fryzlewicz and Subba Rao (2014) if you use this package.

@article{10.1111/rssb.12054,
    author = {Fryzlewicz, P. and Subba Rao, S.},
    title = "{Multiple-Change-Point Detection for Auto-Regressive Conditional Heteroscedastic Processes}",
    journal = {Journal of the Royal Statistical Society Series B: Statistical Methodology},
    volume = {76},
    number = {5},
    pages = {903-924},
    year = {2013},
    month = {11},
    issn = {1369-7412},
    doi = {10.1111/rssb.12054},
    url = {https://doi.org/10.1111/rssb.12054},
    eprint = {https://academic.oup.com/jrsssb/article-pdf/76/5/903/49506908/jrsssb\_76\_5\_903.pdf},
}

Roadmap

At this point, I only implemented stat.resid, bin.segm, inner.prod.iter and med in C++. The rest of the code is mostly unchanged from that of Prof. Piotr Fryzlewicz. I intend to implement more of the code in C++ at some stage.