State estimation of missing data imputation

One intuitive method treating missing values is to use mean imputation. In the absence of better estimates, some forms of mean values are more than often used instead to solve the missing data problem. In spite of heavy criticism in statistical literature, nevertheless, given its simplicity, we attempt to assess the actual impacts from 2D-mean imputations of missing values. This would establish boundaries of its applicability in case of practical necessity. First, we construct a data structure to enable subsequent analysis.