EpiLPS
| REF | Gressani et al. (2022) |
| Docs | |
| Github | github.com/oswaldogressani/EpiLPS |
| Last Commit | Oct, 2024 |
| Installation |
Brief description
Brief summary of the method from the paper
EpiLPS is a Bayesian tool for estimating the time-varying reproduction number using a robust, efficient approach. It models case counts with a Negative Binomial distribution to handle overdispersion and employs Bayesian P-splines for smoothing epidemic curves. The methodology leverages Laplace approximations to estimate the posterior distribution of the spline coefficients rapidly. Two inference methods are provided: a fast maximum a posteriori approach for quick estimates and an MCMC scheme using Langevin dynamics for thorough posterior sampling. EpiLPS delivers accurate estimates without arbitrary smoothing assumptions and has been applied to SARS-CoV-1, H1N1, and COVID-19 datasets.
Methods
This package contains the following methods:
Assessment
| Features | |
| Ability to nowcast/forecast | Nowcasting,adjusts for underreporting by estimating unreported infections and combining them with reported cases to reflect actual daily epidemics |
| Incorporates delay distributions | Some, It accounts for the uncertainty associated with reporting delays |
| Estimates expected cases | Yes |
| Communicates uncertainty | Yes, The credible intervals are calculated via the delta method |
| Validation | |
| Documentation of package methods | Yes |
| Documentation of package implementation | Yes |