Multiple imputation using Fully Conditional Specification (FCS) implemented by the MICE algorithm as described in Van Buuren and Groothuis-Oudshoorn (2011) . Each variable has its own imputation model. Built-in imputation models are provided for continuous data (predictive mean matching, normal), binary data (logistic regression), unordered categorical data (polytomous logistic regression) and ordered categorical data (proportional odds). MICE can also impute continuous two-level data (normal model, pan, second-level variables). Passive imputation can be used to maintain consistency between variables. Various diagnostic plots are available to inspect the quality of the imputations.


Manual: mice.pdf
Vignette: Create simulation data

Maintainer: Stef van Buuren <stef.vanbuuren at>

Author(s): Stef van Buuren*, Karin Groothuis-Oudshoorn*, Alexander Robitzsch*, Gerko Vink*, Lisa Doove*, Shahab Jolani*, Rianne Schouten*, Philipp Gaffert*, Florian Meinfelder*

Install package and any missing dependencies by running this line in your R console:


Depends methods, R (>= 2.10.0)
Imports lattice, grDevices, graphics, MASS, nnet, rpart, splines, stats, survival, utils, Rcpp
Suggests AGD, CALIBERrfimpute, gamlss, lme4, mitools, nlme, pan, randomForest, Zelig, BSDA, knitr, rmarkdown
Linking to Rcpp

Package mice
Task Views Multivariate , OfficialStatistics , SocialSciences
Version 2.30
Published 2017-02-18
License GPL-2 | GPL-3
NeedsCompilation yes
CRAN checks mice check results
Package source mice_2.30.tar.gz