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

View source: R/est.R0.SB.R

est.R0.SBR Documentation

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.

Related to est.R0.SB in R0...