impimp: Imprecise Imputation for Statistical Matching (original) (raw)

Imputing blockwise missing data by imprecise imputation, featuring a domain-based, variable-wise, and case-wise strategy. Furthermore, the estimation of lower and upper bounds for unconditional and conditional probabilities based on the obtained imprecise data is implemented. Additionally, two utility functions are supplied: one to check whether variables in a data set contain set-valued observations; and another to merge two already imprecisely imputed data. The method is described in a technical report by Endres, Fink and Augustin (2018, <doi:10.5282/ubm/epub.42423>).

Documentation:

Downloads:

Linking:

Please use the canonical formhttps://CRAN.R-project.org/package=impimpto link to this page.