EpiBayes (version 0.1.2)

summary.ebhistorical: Summary Method for EpiBayes Historical Object

Description

This function gives summary measurements for posterior distributions of cluster-level prevalences across all time periods considered. It does so by examining the object output by the EpiBayesHistorical function of class ebhistorical.

Usage

"summary"(object, sumstat = "mean", prob = 0.95, time.labels = NULL, busterapprox = FALSE, burnin = NULL, ...)

Arguments

object

An object of class ebhistorical (e.g., the output of function EpiBayesHistorical).

sumstat

The summary statistic that the user wishes to be output. The three choices are mean, quantile, and variance. Character scalar.

prob

The probability associated with the quantile one wishes to calculate. Only used if sumstat is chosen to be quantile. Real scalar.

time.labels

Labels for the time period axis (e.g., a vector of years). Character vector.

busterapprox

Boolean value indicating whether or not the summary statistics should be computed using the raw posterior distribution values computed via MCMC (TRUE) or using the best beta distribution approximation via moment matching (FALSE). Boolean.

burnin

Number of MCMC iterations to discard from the beginning of the chain. Integer scalar.

...

Additional arguments to be passed on to summary.

Value

The summary statistics are returned in a matrix with rows indexed by time period and columns indexed by subzone. The body of the matrix is filled with the summary statistics requested for by the argument sumstat.

See Also

This is a method for objects of class ebhistorical returned by the function EpiBayesHistorical and creates its own class of object much like the summary method for lm objects. There is an additional plot method that will plot the summary output from this method, summary.ebhistorical.