miceadds (version 1.5-0)

mice.impute.grouped: Imputation of a Variable with Grouped Values

Description

Imputes a variable with continuous values whose original values are only available as grouped values.

Usage

mice.impute.grouped(y, ry, x, low=NULL , upp=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.

low

Vector with lower bound of grouping interval

upp

Vector with upper bound of grouping interval

...

Further arguments to be passed

Value

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

See Also

This function uses the grouped package.

Examples

#############################################################################
# EXAMPLE 1: Imputation of grouped data
#############################################################################	
	
data(data.ma06)
data <- data.ma06

# define the variable "FC_imp" which should contain the variables to be imputed
data$FC_imp <- NA
V <- ncol(data)
# variables not to be used for imputation 
vars_elim <-  c("id" , "FC","FC_low","FC_upp")

# define imputation methods
impM <- rep("norm" , V)
names(impM) <- colnames(data)
impM[  vars_elim ] <- ""
impM[ "FC_imp" ] <- "grouped"

# define predictor matrix
predM <- 1 - diag( 0 , V)
rownames(predM) <- colnames(predM) <- colnames(data)
predM[vars_elim, ] <- 0
predM[,vars_elim] <- 0

# define lower and upper boundaries of the grouping intervals
low <- list("FC_imp" = data$FC_low )
upp <- list("FC_imp" = data$FC_upp )

# perform imputation
imp <- mice( data , imputationMethod = impM , predictorMatrix = predM , 
	    m=1 , maxit=3 , allow.na=TRUE ,  low=low , upp=upp)
head( complete(imp ))