midastouch_donors: Midastouch: PMM with covariate-distance-weighted donor...

In VIM: Visualization and Imputation of Missing Values

Midastouch: PMM with covariate-distance-weighted donor selection

Description

For each missing value, finds k nearest donors using a combined score: closeness in predicted value (score) AND closeness in covariate space (Mahalanobis distance). Donors closer in covariate space are upweighted.

Usage

midastouch_donors(
  y_obs,
  X_obs,
  X_miss,
  score_obs = NULL,
  score_miss = NULL,
  k = 5L
)

Arguments

y_obs	Observed values of the target variable
X_obs	Predictor matrix for observed rows (n_obs x p)
X_miss	Predictor matrix for missing rows (n_miss x p)
score_obs	Model predictions for observed rows
score_miss	Model predictions for missing rows
k	Number of candidate donors (default 5)

Details

Based on Siddique & Belin (2008), "Multiple imputation using an iterative hot-deck with distance-based donor selection", Statistics in Medicine.

Value

Numeric vector of length n_miss with imputed values drawn from donors

VIM documentation built on Sept. 2, 2026, 5:07 p.m.