EpiEstim
| REF | Cori et al. (2013), Nash et al. (2023) |
| Docs | mrc-ide.github.io/EpiEstim |
| Github | github.com/mrc-ide/EpiEstim |
| Last commit | Aug 30, 2024 |
| Installation | CRAN |
Brief description
EpiEstim is a tool to estimate the time-varying instantaneous reproduction number during epidemics. In order to estimate Rt, EpiEstim needs to be supplied with an estimate of the serial interval distribution (step A) and the incidence of confirmed cases (step B).
Once you have an incidence object (based on the dates of symptom onset) and information on the serial interval distribution, we can use the renewal equation (a form of branching process model) to estimate Rt. The incidence of symptom onset at time t is approximated by a Poisson process using the renewal equation.
Assuming some prior distributions for \(R(t)\) and the generation interval permit an analytical solution for the posterior distribution of \(R(t)\), as in EpiEstim. These simplifying assumptions greatly constrain the space of possible \(R(t)\) values and thus calculation times are relatively fast.
Methods
This package contains the following methods:
Assessment
| Features | |
| Ability to nowcast/forecast | No |
| Incorporates delay distributions | No, although some right-censoring is included |
| Estimates expected cases | No |
| Communicates uncertainty | Yes |
| Validation | |
| Documentation of package methods | Yes |
| Documentation of package implementation | Yes |
Sample Code
This vignette gives a basic example of usage of EpiEstim.
The end of this vignette suggests using the projections package to estimate future cases, and we cannot recommend this package. The estimation of future values of R(t) in this package comes from resampling different past values of R(t) rather than trends dervied from recent infections.