arima2: Likelihood Based Inference for ARIMA Modeling

Estimating and analyzing auto regressive integrated moving average (ARIMA) models. The primary function in this package is arima(), which fits an ARIMA model to univariate time series data using a random restart algorithm. This approach frequently leads to models that have model likelihood greater than or equal to that of the likelihood obtained by fitting the same model using the arima() function from the 'stats' package. This package enables proper optimization of model likelihoods, which is a necessary condition for performing likelihood ratio tests. This package relies heavily on the source code of the arima() function of the 'stats' package. For more information, please see Jesse Wheeler and Edward L. Ionides (2023) <arXiv:2310.01198>.

Version: 3.1.0
Depends: R (≥ 3.5)
Imports: ggplot2, methods
Suggests: testthat (≥ 3.0.0)
Published: 2023-11-16
Author: Jesse Wheeler [aut, cre, cph], Noel McAllister [aut], Dhajanae Sylvertooth [aut], Edward Ionides [ctb], Brian Ripley [ctb] (Author of arima source code in stats package.), R Core Team [cph] (Author of arima source code in stats package.)
Maintainer: Jesse Wheeler <jeswheel at>
License: GPL (≥ 3)
NeedsCompilation: yes
Materials: README NEWS
In views: TimeSeries
CRAN checks: arima2 results


Reference manual: arima2.pdf


Package source: arima2_3.1.0.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
macOS binaries: r-release (arm64): arima2_3.1.0.tgz, r-oldrel (arm64): arima2_3.1.0.tgz, r-release (x86_64): arima2_3.1.0.tgz, r-oldrel (x86_64): arima2_3.0.1.tgz
Old sources: arima2 archive


Please use the canonical form to link to this page.