Available on CRAN and R-universe
An R library for paleoecology and regime shift analysis.
These functions assume your data is in tidy format.
rshift leverages extendr to use algorithms written in Rust, making it faster
than any R implementation!
rshift is available on CRAN and R-universe, so can be installed via the RStudio package manager, or by running
install.packages('rshift', repos = c('https://alexhroom.r-universe.dev', 'https://cloud.r-project.org'))
Note that as rshift uses Rust code, if building from source rustc and cargo are required.
rshift mainly focuses around Rodionov (2004)'s STARS algorithm. A detailed explanation of the algorithm,
and how to use it for regime shift analysis in R, is available here,
or in R itself with the command vignette("STARSmanual")
It also contains a few other tools for analysing time series/paleoecological data.
For further documentation and help with rshift, either run help(packages = "rshift"),
or view the documentation here.
rshift was originally developed as part of a NERC GW4+ undergraduate placement,
with Dunia H. Urrego and Felipe Franco-Gaviria of the University of Exeter.
In particular, it was part of the BioResilience project.
Now, it is maintained and improved on a volunteer/free time basis; if you have any improvements or suggestions, please create an issue or pull request as necessary.
(If you use rshift in your publication, please let me know and I'd love to include it here!)