A Bayesian hierarchical model with integrated covariate selection and misclassification matrices to estimate neonatal causes of death

Code and functions for paper by Amy MULICK, Shefali OZA, David PRIETO-MERINO, Francisco VILLAVICENCIO, Simon COUSENS and Jamie PERIN, published at the Journal of the Royal Statistical Society, Series A, 24 June 2022. Full text available at: http://doi.org/10.1111/rssa.12853

This repository contains the following files and folders:

Bayesian_COD.R: This file can be executed to upload data and run the analysis for the final Bayesian cause of death model from the published paper. Any reader should be able to run the code in R or RStudio, with no need to manipulate the file, but installing/uploading all the necessary libraries.

The Data subfolder contains all the necessary data to reproduce the final model and out-of-sample predictions. The data will be automatically uploaded when executing the R-script.