mice.impute.hotDeck.RdImputes a variable under a random draw from a pool of donors defined by a distance function. Uncertainty with respect to the creation of donor pools is introduced by drawing a Bootstrap sample (approximate Bayesian Bootstrap, ABB) from observations with complete data (see Andridge & Little, 2010).
mice.impute.hotDeck(y, ry, x, donors=5, method="Mahalanobis", ...)Incomplete data vector of length n
Vector of missing data pattern (FALSE -- missing,
TRUE -- observed)
Matrix (n x p) of complete covariates.
Number of donors used for random sampling of nearest neighbors in imputation
Method used for computation of weights in distance function.
Options are the Mahalanobis metric (method="Mahalanobis"),
weighted by correlations of covariates with the outcome
(method="cor") and weighting by linear regression
coefficients (method="lm").
Further arguments to be passed
A vector of length nmis=sum(!ry) with imputed values.
Andridge, R. R., & and Little, R. J. A. (2010). A review of hot deck imputation for survey non-response. International Statistical Review, 78(1), 40-64. doi:10.1111/j.1751-5823.2010.00103.x
See also the packages hot.deck and HotDeckImputation.
if (FALSE) {
#############################################################################
# EXAMPLE 1: Hot deck imputation NHANES dataset
#############################################################################
data(nhanes, package="mice")
dat <- nhanes
#*** prepare imputation method
vars <- colnames(dat)
V <- length(vars)
impMethod <- rep("hotDeck", V)
method <- "cor"
#*** imputation in mice
imp <- mice::mice( data=as.matrix(dat), m=1, method=impMethod, method=method )
summary(imp)
}