est.R0.SB: Estimate the time dependent R using a Bayesian method

In R0: Estimation of R0 and Real-Time Reproduction Number from Epidemics

Estimate the time dependent R using a Bayesian method

Description

Estimate R by a sequential Bayesian method, using known data up to a point in time as a Bayesian prior for the next iteration (see details).

Usage

est.R0.SB(
  epid,
  GT,
  t = NULL,
  begin = NULL,
  end = NULL,
  date.first.obs = NULL,
  time.step = 1,
  force.prior = FALSE,
  checked = FALSE,
  ...
)

Arguments

epid	Object containing epidemic curve data.
GT	Generation time distribution from generation.time().
t	Vector of dates at which incidence was observed.
begin	At what time estimation begins (unused by this method, just there for plotting purposes).
end	At what time estimation ends (unused by this method, just there for plotting purposes).
date.first.obs	Optional date of first observation, if t not specified.
time.step	Optional. If date of first observation is specified, number of day between each incidence observation.
force.prior	Set to any custom value to force the initial prior as a uniform distribution on [0 ; value].
checked	Internal flag used to check whether integrity checks were ran or not.
...	Parameters passed to inner functions.

Details

For internal use. Called by estimate.R().

Initial prior is an unbiased uniform distribution for R, between 0 and the maximum of incid(t+1) - incid(t). For each subsequent iteration, a new distribution is computed for R, using the previous output as new prior.

The 95% confidence intervan is achieved by a cumulative sum of the posterior distirbution of R and corresponds to the 2.5-th and 97.5-th percentiles.

Value

A list with components:

R	vector of R values.
conf.int	95% confidence interval for estimates.
proba.Rt	A list with successive distribution for R throughout the outbreak.
GT	Generation time distribution used in the computation.
epid	Original epidemic data.
begin	Begin date for the fit.
begin.nb	Index of begin date for the fit.
end	End date for the fit.
end.nb	Index of end date for the fit.
pred	Predictive curve based on most-likely R value.
data.name	Name of the data used in the fit.
call	Complete call used to generate results.
method	Method for estimation.
method.code	Internal code used to designate method.

Note

This is the implementation of the method provided by Bettencourt & Ribeiro (2008).

Author(s)

Pierre-Yves Boelle, Thomas Obadia

References

Bettencourt, L.M.A., and R.M. Ribeiro. "Real Time Bayesian Estimation of the Epidemic Potential of Emerging Infectious Diseases." PLoS One 3, no. 5 (2008): e2185.

Examples

#Loading package
library(R0)

## Data is taken from the paper by Nishiura for key transmission parameters of an institutional
## outbreak during 1918 influenza pandemic in Germany)

data(Germany.1918)
mGT <- generation.time("gamma", c(3,1.5))
SB <- est.R0.SB(Germany.1918, mGT)

## Results will include "most likely R(t)" (ie. the R(t) value for which the computed probability 
## is the highest), along with 95% CI, in a data.frame object
SB
# Reproduction number estimate using  Real Time Bayesian  method.
# 0 0 2.02 0.71 1.17 1.7 1.36 1.53 1.28 1.43 ...

SB$Rt.quant
# Date R.t. CI.lower. CI.upper.
# 1  1918-09-29 0.00      0.01      1.44
# 2  1918-09-30 0.00      0.01      1.42
# 3  1918-10-01 2.02      0.97      2.88
# 4  1918-10-02 0.71      0.07      1.51
# 5  1918-10-03 1.17      0.40      1.84
# 6  1918-10-04 1.70      1.09      2.24
# 7  1918-10-05 1.36      0.84      1.83
# 8  1918-10-06 1.53      1.08      1.94
# 9  1918-10-07 1.28      0.88      1.66
# 10 1918-10-08 1.43      1.08      1.77
# ...

## "Plot" will provide the most-likely R value at each time unit, along with 95CI
plot(SB)
## "Plotfit" will show the complete distribution of R for 9 time unit throughout the outbreak
plotfit(SB)

R0 documentation built on Sept. 26, 2023, 5:10 p.m.