Missing Data Imputation Using Gaussian Copulas
Get mdgc Object
Get Pointer to C++ Object to Approximate the Log Marginal Likelihood
mdgc: Missing Data imputation using Gaussian Copulas
Perform Model Estimation and Imputation
Estimate the Model Parameters
Impute Missing Values
Evaluate the Log Marginal Likelihood and Its Derivatives
Get Starting Value for the Covariance Matrix Using a Heuristic
Provides functions to impute missing values using Gaussian copulas for mixed data types as described by Christoffersen et al. (2021) <arXiv:2102.02642>. The method is related to Hoff (2007) <doi:10.1214/07-AOAS107> and Zhao and Udell (2019) <arXiv:1910.12845> but differs by making a direct approximation of the log marginal likelihood using an extended version of the Fortran code created by Genz and Bretz (2002) <doi:10.1198/106186002394> in addition to also support multinomial variables.