Online Program

Official statistics and survey methodology meets R: An overview of corresponding packages

Keywords:

The free and open-source programming language and software environment R is currently both, the most widely used and most popular software for statistics and data analysis.

The basic R environment can be obtained from the comprehensive R archive network (CRAN, http://cran.r-project.org). R is extendible via packages, which consist of code and structured standard documentation including code application examples and possible further documents (so called vignettes) that showing further applications of the packages and dependencies between the implemented functions and methods.

The CRAN Task View on Official Statistics and Survey Methodology (http://CRAN.R-project.org/view=OfficialStatistics) gives a comprehensive overview of those packages which are useful in that area, currently categorized in the topics: Complex Survey Design including point and variance estimation and calibration, editing, imputation, statistical disclosure control, seasonal adjustment, statistical record matching, indices and indicators. In this contribution an overview of all important packages in those areas are given.

To demonstrate the usefulness of those packages to real-world problems few applications of some specific packages will be additionally shown by using, for example, package - VIM: visualisation and imputation of missing values) - simPopulation: generation of synthetic data - sdcMicro and sdcTable: both for statistical disclosure control - simFrame: for simulation studies in official statistics - rrcovNA: multivariate outlier detection of data sets with missing values) - sparkTable: visualising graphical tables and/or - laeken: robust semi-parametric estimation of indicators

Finally, the possibilities of using R in the statistical office regarding data import/export, data base connections, data formats and dynamical reporting are briefly outlined.