Provides robust methods to detect change-points in uni- or multivariate time series. They can cope with corrupted data and heavy tails. One can detect changes in location, scale and dependence structure of a possibly multivariate time series. Procedures are based on Huberized versions of CUSUM tests proposed in Duerre and Fried (2019) <arXiv:1905.06201>.
| Version: | 0.2.4 |
| Depends: | R (≥ 3.3.1) |
| Published: | 2019-06-03 |
| Author: | Sheila Goerz [aut, cre], Alexander Duerre [ctb] |
| Maintainer: | Sheila Goerz <sheila.goerz at tu-dortmund.de> |
| License: | GPL-3 |
| NeedsCompilation: | yes |
| In views: | TimeSeries |
| CRAN checks: | robcp results |
| Reference manual: | robcp.pdf |
| Package source: | robcp_0.2.4.tar.gz |
| Windows binaries: | r-devel: robcp_0.2.4.zip, r-release: robcp_0.2.4.zip, r-oldrel: robcp_0.2.4.zip |
| OS X binaries: | r-release: robcp_0.2.4.tgz, r-oldrel: robcp_0.2.4.tgz |
Please use the canonical form https://CRAN.R-project.org/package=robcp to link to this page.