Multiple Imputation Using MICE and Random Forest

Functions to impute using random forest under full conditional specifications (multivariate imputation by chained equations). The methods are described in Shah and others (2014) .


News

===================================== Version 1.0-1 sixth release on CRAN

BUG FIXES:

  • Vignette updated to work with mice 3 format of multiple imputation results.

===================================== Version 1.0-0 fifth release on CRAN

CHANGES:

  • Allow mice.impute.rfcat to work with logical variables (consistent with mice.impute.logreg)

===================================== Version 0.1-6 fourth release on CRAN

BUG FIXES:

  • Errors in mice.impute.rfcat and mice.impute.rfcont when there is one missing value and multiple predictors (thanks to David Erceg-Hurn for finding this bug)
  • Error in mice.impute.rfcat when all observed values are identical (thanks to David Erceg-Hurn for finding this bug)

===================================== Version 0.1-5 third release on CRAN

CHANGES:

  • Update vignette to investigate RF Doove with different numbers of trees. Remove rfImpute from vignette, as it is not suitable for survival data
  • Updated author names in DESCRIPTION.

BUG FIXES:

  • Corrections to vignette so that it works on multiple platforms
  • Ensure that rfcont works with a single column predictor - randomForest requires the predictor to be a matrix even if one column (thanks to Jonathan Bartlett for finding this bug)

===================================== Version 0.1-4 second release on CRAN

CHANGES:

  • A new argument ntree, to correspond to the ntree argument in mice.impute.rf

===================================== Version 0.1-3

CHANGES:

  • Added information and links for the mice.impute.rf and mice.impute.cart methods in mice version 2.20
  • Added reference to published paper

===================================== Version 0.1-2 first release to CRAN

CHANGES:

  • A new option: maxnodes (maximum number of nodes in regresion tree)

===================================== Version 0.1-1 second version on R-Forge

BUG FIXES:

  • Allow factors with missing categories in rfcat
  • Allow single predictor for imputation models

===================================== Version 0.1-0

Initial version

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("CALIBERrfimpute")

1.0-6 by Anoop Shah, a day ago


Browse source code at https://github.com/cran/CALIBERrfimpute


Authors: Anoop Shah [aut, cre] , Jonathan Bartlett [ctb] , Harry Hemingway [ths] , Owen Nicholas [ths] , Aroon Hingorani [ths]


Documentation:   PDF Manual  


Task views:


GPL-3 license


Imports mvtnorm, randomForest

Depends on mice

Suggests missForest, rpart, survival, xtable


See at CRAN