![[Translate to Englisch:] Ein Arbeitsplatz mit Bildschirm und Tastatur, jemand tippt gerade und im Hintergrund sieht man eine andere Person die gerade auf ein Dokument deutet](/fileadmin/_processed_/5/0/csm_R-Pakete_gross_5aa9954644.jpg)
Objective and implementation
Today most of the tasks of Statistics Austria’s methods unit are covered by functions implemented in (in-house developed) R packages. Some of them are developed for internal use only, however, a number of packages are also made available publicly. Statistics Austria is steadily encouraging the usage of R internally, but is also actively contributing to the international open source community by further developing R packages relevant for official statistics.
Innovation within the project
In academia, R is considered the lingua franca for statistics with regard to advancing statistical methods and tools and their implementation. Statistics Austria is therefore continuously promoting the usage of R and thus the exchange with academia and other NSIs. The packages developed at Statistics Austria are applicable to statistical production processes of NSIs around the world.
Further information, project results
Github (https://github.com/statistikat) is the development platform for externally available R packages. All R packages that are in regular production are published on CRAN (https://cran.r-project.org/) as well. Some of the packages are even listed on the list "Awesome official statistics software" (http://www.awesomeofficialstatistics.org ).
Developed R packages
Statistical disclosure control (SDC):
- sdcMicro (Templ et al., 2015) SDC of micro data
- sdcTable (Meindl, 2017) SDC of tabular data
- simPop (Temp et al., 2017) model based generation of synthetic micro data (mainly for public use files)
Imputation:
- VIM (Kowarik et al., 2016) Visualization and Imputation of missing data
Seasonal adjustment:
- x12 (Kowarik et al., 2014) functionality to use X13-ARIMA-SEATS directly from within R
- persephone (https://github.com/statistikat/persephone) a wrapper for JDemetra+, the official seasonal adjustment tool of the ESS. This package is still under development and is planned to completely replace the package x12 in the future.
Weighting and error estimation (Bootstrap) of sample surveys:
- surveysd (Gussenbauer et al., 2020) implements a bootstrap for the error estimation for complex survey designs and a flexible calibration method "Iterative proportional fitting".
Publications:
- Gussenbauer J., Kowarik A, de Cillia G., 2020, surveysd: Survey Standard Error Estimation for Cumulated Estimates and their Differences in Complex Panel Designs, R package version 1.2.0, https://CRAN.R-project.org/package=surveysd
- Kowarik, A., & Templ, M., 2016, Imputation with R package VIM. Journal of Statistical Software, 74(7), 1-16.
- Kowarik, A., Meraner, A., Templ, M., & Schopfhauser, D., 2014, Seasonal Adjustment with the R Packages x12 and x12GUI. Journal of Statistical Software, 62(1),
1-21. - Meindl, B., 2017, sdcTable: Methods for SDC (statistical disclosure control) in
tabular data, https://cran.r-project.org/web/packages/sdcTable/index.html. - Templ, M., Meindl, B., Kowarik, A., & Dupriez, O., 2017, Simulation of synthetic
complex data: The R-package simPop. Journal of Statistical Software, 79(i10). - Templ, M., Kowarik, A., & Meindl, B., 2015, Statistical disclosure control for microdata using the R package sdcMicro. Journal of Statistical Software, 67(1), 1-36
Access to open data:
-
STATcubeR (https://statistikat.github.io/STATcubeR/) R Client for STATcube REST API and open data portal (data.statistik.gv.at)