fit.epid: Poisson log-likelihood for an observed epidemic
In R0: Estimation of R0 and Real-Time Reproduction Number from Epidemics

fit.epid

R Documentation

Poisson log-likelihood for an observed epidemic

Description

Computes the Poisson log-likelihood of an observed epidemic, compared to the expected incidence values for a given R and generation time distirbution.

Usage

fit.epid(log.R, epid, GT, import, pred = FALSE, offset = 0)

Arguments

`log.R`	The log-reproduction ratio.
`epid`	An epidemic curve, in the sense of passing `check.incid()` validations. It must have at least an incidence vector in an `incid` component.
`GT`	Generation time distribution from `generation.time()`.
`import`	Vector of imported cases
`pred`	Boolean. By default (`FALSE`), the function returns the corresponding log-likelihood. When set to `TRUE`, the function will return the Poisson-predicted incidence series instead.
`offset`	Offset value for the log-likelihood (used to calculate confidence intervals).

Details

For internal use. Called from est.R0.ML().

Computes and returns the Poisson log-likelihood of an observed epidemic, conditional on a given value of R and a generation time distribution.

Value

The Poisson log-likelihood of the observed epidemic passed as argument epid, or the Poisson-predicted incidence series given R and GT values.

Note

This function is also used in fit.epid.optim(), which is a wrapper for the joint optimization routine to determine the best-fitting R and GT values when the est.R0.ML() method is called and asked to jointly estimate reproduction ratios and generation time distirbution.