StanMoMo package performs Bayesian Mortality Modeling with Stan, which is a C++ package for performing full Bayesian inference (see https://mc-stan.org/). The current package supports a variety of popular mortality models: the Lee-Carter (LC) model, the Renshaw-Haberman model (LC with cohort effect), the Age-Period-Cohort (APC) model, the Cairns-Blake-Dowd (CBD) model and the M6 model (CBD with cohort effect). By a simple function call, the user obtains the MCMC simulations for each parameter, the log likelihoods and death rates predictions. Moreover, the package includes tools for model selection and Bayesian model averaging by leave-future-out validation.
To install the stable version on CRAN:
To install the source package from Github:
The installation of the source package may take a few minutes (models need to be compiled). For this reason, we recommend to install the binary package instead. Once the package is installed, you can perform Bayesian mortality forecasting in a matter of seconds.
After installing the package, you have to load the package via:
If you have any comments, suggestions for improvements or if you are motivated to contribute to the package, feel free to email email@example.com
This package is related to our paper. When referring to this package, please cite as
Barigou, K., Goffard, P. O., Loisel, S., & Salhi, Y. (2023). Bayesian model averaging for mortality forecasting using leave-future-out validation. International Journal of Forecasting, 39(2), 674-690. https://doi.org/10.1016/j.ijforecast.2022.01.011