Consistent Perturbation of Statistical Frequency- And Magnitude Tables

Data from statistical agencies and other institutions often need to be protected before they can be published. This package can be used to perturb statistical tables in a consistent way. The main idea is to add - at the micro data level - a record key for each unit. Based on these keys, for any cell in a statistical table a cell key is computed as a function on the record keys contributing to a specific cell. Values that are added to the cell in order to perturb it are derived from a lookup-table that maps values of cell keys to specific perturbation values. The theoretical basis for the methods implemented can be found in Thompson, Broadfoot and Elazar (2013) < https://unece.org/fileadmin/DAM/stats/documents/ece/ces/ge.46/2013/Topic_1_ABS.pdf> which was extended and enhanced by Giessing and Tent (2019) < https://unece.org/fileadmin/DAM/stats/documents/ece/ces/ge.46/2019/mtg1/SDC2019_S2_Germany_Giessing_Tent_AD.pdf>.

cellKey

an R package for applying noise to statistical tables.

Information

This package is developed within the SGA Open source tools for perturbative confidentiality methods. This package is not fully optimized for speed but already contains the core functionality to compute perturbed count- and magnitude tables with complex hierarchies.

Feedback (via issues) with regards to bugs or features requests are very welcome as well as pull-requests.

Vignette

The core-concepts of the package and its application is described in the package vignette that can be viewed after installation of the package using ck_vignette() or online.

Installation

The package can directly be installed from CRAN using the install.packages() function which automatically also installs the required dependencies (such as ptable.

# update your R installation
install.packages("cellKey")

Usage

An example on how to apply the package is provided in ?cellKey::cellkey_pkg where also all the available methods are described.

?cellKey::cellkey_pkg

The package vignette can be looked at using cellKey::ck_vignette() or via the automatically deployed documentation by clicking here. The complete documentation is also updated automatically and can be viewed online.