Detects multiple changepoints in uni- or multivariate time series data. The algorithm is based on Bayesian methods and detects changes on-line; i.e. the model updates with every observation rather than relying on retrospective segmentation. However, the user may choose to use the algorithm off- or on-line.
| Version: | 1.0 |
| Depends: | scales, gridExtra, reshape2 |
| Imports: | ggplot2 |
| Published: | 2016-08-23 |
| Author: | Zachary Zanussi |
| Maintainer: | Zachary Zanussi <zachary9506 at gmail.com> |
| License: | GPL-3 |
| NeedsCompilation: | no |
| In views: | TimeSeries |
| CRAN checks: | onlineCPD results |
| Reference manual: | onlineCPD.pdf |
| Package source: | onlineCPD_1.0.tar.gz |
| Windows binaries: | r-devel: onlineCPD_1.0.zip, r-release: onlineCPD_1.0.zip, r-oldrel: onlineCPD_1.0.zip |
| OS X Mavericks binaries: | r-release: onlineCPD_1.0.tgz, r-oldrel: onlineCPD_1.0.tgz |
Please use the canonical form https://CRAN.R-project.org/package=onlineCPD to link to this page.