tsrobprep: Robust Preprocessing of Time Series Data

Methods for handling the missing values outliers are introduced in this package. The recognized missing values and outliers are replaced using a model-based approach. The model may consist of both autoregressive components and external regressors. The methods work robust and efficient, and they are fully tunable. The primary motivation for writing the package was preprocessing of the energy systems data, e.g. power plant production time series, but the package could be used with any time series data. For details, see Narajewski et al. (2021) <doi:10.1016/j.softx.2021.100809>.

Version: 0.3.2
Depends: R (≥ 3.2.0)
Imports: glmnet, MASS, Matrix, mclust, quantreg, Rdpack, splines, textTinyR, zoo
Published: 2022-02-22
Author: Michał Narajewski ORCID iD [aut, cre], Jens Kley-Holsteg [aut], Florian Ziel ORCID iD [aut]
Maintainer: Michał Narajewski <michal.narajewski at uni-due.de>
License: MIT + file LICENSE
NeedsCompilation: no
Citation: tsrobprep citation info
In views: MissingData, TimeSeries
CRAN checks: tsrobprep results

Documentation:

Reference manual: tsrobprep.pdf

Downloads:

Package source: tsrobprep_0.3.2.tar.gz
Windows binaries: r-devel: tsrobprep_0.3.2.zip, r-release: tsrobprep_0.3.2.zip, r-oldrel: tsrobprep_0.3.2.zip
macOS binaries: r-release (arm64): tsrobprep_0.3.2.tgz, r-oldrel (arm64): tsrobprep_0.3.2.tgz, r-release (x86_64): tsrobprep_0.3.2.tgz, r-oldrel (x86_64): tsrobprep_0.3.2.tgz
Old sources: tsrobprep archive

Linking:

Please use the canonical form https://CRAN.R-project.org/package=tsrobprep to link to this page.