Guide for endemic-epidemic models in RTMB

Author

Sophie Reichert



Preface

This guide describes how endemic-epidemic models from the R package surveillance can be implemented using the R package RTMB. Each chapter compares hhh4 models and some extensions with the corresponding RTMB implementations. We demonstrate what is required to implement multivariate time-series endemic-epidemic models, and explore some options offered by RTMB.

For each model, we compare the parameter estimates, (negative) log-likelihoods, and computational runtime of both implementations. The aim is to demonstrate that the results of hhh4 models can be reproduced using the RTMB package, which potentially provides computational advantages through its automatic differentiation approach and more flexibility for further extensions of endemic-epidemic models.

License

This work is licensed under Creative Commons Attribution 4.0 International