Gravar-mail: Fully Bayesian inference under ignorable missingness in the presence of auxiliary covariates