miceadds (version 1.5-0)

Rhat.mice: Rhat Convergence Statistic of a mice Imputation

Description

Computes the Rhat statistic for a mids object.

Usage

Rhat.mice(mice.object)

Arguments

mice.object

Object of class mids

Value

• Data frame containing the Rhat statistic for mean and variances for all variables of the Markov chains used for imputation

References

Gelman, A., & Hill, J. (2007). Data analysis using regression and multilevel/hierarchical models. Cambridge University Press.

Examples

#############################################################################
# EXAMPLE 1: Rhat statistic for nhanes data 
#############################################################################

library(mice)
data(nhanes, package="mice")
set.seed(9090)

# nhanes 3 parallel chains
imp1 <- mice( nhanes ,m=3 , maxit=10 , imputationMethod=rep("norm" , 4 ) )
Rhat.mice( imp1 )
  ##     variable MissProp Rhat.M.imp Rhat.Var.imp
  ##   1      bmi       36  1.0181998     1.155807
  ##   2      hyp       32  1.0717677     1.061174
  ##   3      chl       40  0.9717109     1.318721