milonga: Multiple imputation for multivariate binary outcome

In milonga: Multiple Imputation for Multivariate Binary Outcome

Description

Multiple imputation for multivariate binary outcome by Gibbs' sampling of all potential profiles.

Usage

milonga(data, null_p=NULL, weight=10, burnin=NULL, rep=10) 

Arguments

data	Data frame or matrix of a multivariate binary outcome with missing values to impute, NA is for missing values
null_p	Prespecified probability of the all-zero profile. The default is NULL where 1/2^T is applied
weight	A weight between study population for observed frequencies and a weighted population for unobserved frequencies, default is 10
burnin	Number of burnin iterations, default is NULL
rep	Number of imputation iterations, default is 10

Details

TBA

Value

burnin a list of burnin datasets if burn is specified.

imp a list of imputation datasets

Examples

data(polio)
null_p<-0.8
out<-milonga(polio, null_p, burnin=5)
names(out)

milonga documentation built on Jan. 15, 2017, 9:47 p.m.