Imputation by Weighted Predictive Mean Matching or Weighted Normal Linear Regression

Imputation by predictive mean matching or normal linear regression using sampling weights.

mice.impute.weighted.pmm(y, ry, x, wy=NULL, imputationWeights=NULL,
      pls.facs=NULL, interactions=NULL, quadratics=NULL, donors=5, ...)

mice.impute.weighted.norm(y, ry, x, wy=NULL, ridge=1e-05, pls.facs=NULL,
     imputationWeights=NULL, interactions=NULL, quadratics=NULL, ...)

Arguments

y: Incomplete data vector of length n
ry: Vector of missing data pattern (FALSE -- missing, TRUE -- observed)
x: Matrix (n x p) of complete covariates.
wy: Logical vector of length length(y). A TRUE value indicates locations in y for which imputations are created.
imputationWeights: Optional vector of sampling weights
pls.facs: Number of factors in PLS regression (if used). The default is NULL which means that no PLS regression is used for dimension reduction.
interactions: Optional vector of variables for which interactions should be created
quadratics: Optional vector of variables which should also be included as quadratic effects.
donors: Number of donors
...: Further arguments to be passed
ridge: Ridge parameter in the diagonal of \( \bold{X}'\bold{X}\)

Value

A vector of length nmis=sum(!ry) with imputed values.

Examples

if (FALSE) {
#############################################################################
# EXAMPLE 1: Imputation using sample weights
#############################################################################

data( data.ma01)
set.seed(977)

# select subsample
dat <- as.matrix(data.ma01)
dat <- dat[ 1:1000, ]

# empty imputation
imp0 <- mice::mice( dat, maxit=0)

# redefine imputation methods
meth <- imp0$method
meth[ meth=="pmm"  ] <- "weighted.pmm"
meth[ c("paredu", "books", "migrant" ) ] <- "weighted.norm"
# redefine predictor matrix
pm <- imp0$predictorMatrix
pm[, 1:3 ] <- 0
# do imputation
imp <- mice::mice( dat, predictorMatrix=pm, method=meth,
           imputationWeights=dat[,"studwgt"], m=3, maxit=5)
}