PMMI: Predictive Mean Matching with Multiple Imputation

In DLMRMV: Distributed Linear Regression Models with Response Missing Variables

Predictive Mean Matching with Multiple Imputation

Description

Implements PMM algorithm for handling missing data in linear regression models. Uses chained equations approach to generate multiple imputed datasets and pools results using Rubin's rules.

Usage

PMMI(data, k = 5, m = 5)

Arguments

data	Dataframe with response variable in 1st column and predictors in others
k	Number of nearest neighbors for matching (default=5)
m	Number of imputations (default=5)

Value

List containing:

Y	Original response vector with NAs
Yhat	Final imputed response vector (averaged across imputations)
betahat	Pooled regression coefficients
imputations	List of m completed datasets
m	Number of imputations performed
k	Number of neighbors used

Examples

# Create dataset with 30% missing values
data <- data.frame(Y=c(rnorm(70),rep(NA,30)), X1=rnorm(100))
results <- PMMI(data, k=5, m=5)

DLMRMV documentation built on Aug. 8, 2025, 6:27 p.m.