interval_credible: Bayesian Credible Interval

In ArchaeoStat/ArchaeoPhases: Post-Processing of Markov Chain Monte Carlo Simulations for Chronological Modelling

Bayesian Credible Interval

Description

Computes the shortest credible interval of the output of the MCMC algorithm for a single parameter.

Usage

interval_credible(x, ...)

## S4 method for signature 'MCMC'
interval_credible(
  x,
  level = 0.95,
  calendar = getOption("ArchaeoPhases.calendar")
)

Arguments

x	An MCMC object containing the output of the MCMC algorithm.
...	Currently not used.
level	A length-one numeric vector giving the confidence level.
calendar	A aion::TimeScale object specifying the target calendar (see calendar()).

Details

A (100 \times level) % credible interval is an interval that keeps N \times (1 - level) elements of the sample outside the interval.

The (100 \times level) % credible interval is the shortest of all those intervals.

For instance, the 95% credible interval is the central portion of the posterior distribution that contains 95% of the values.

Value

Returns a list of numeric matrix.

Author(s)

A. Philippe, M.-A. Vibet, T. S. Dye, N. Frerebeau

See Also

arkhe::interval_credible()

Other statistics: interval_hdr(), sensitivity(), summary()

Examples

## Coerce to events
eve <- as_events(mcmc_events, calendar = CE(), iteration = 1)
eve <- eve[1:10000, ]

## Rata die
interval_credible(eve, level = 0.95) # Credible interval
interval_hdr(eve, level = 0.68) # HPD interval

## BP
interval_credible(eve, level = 0.95, calendar = BP()) # Credible interval
interval_hdr(eve, level = 0.95, calendar = BP()) # HPD interval

ArchaeoStat/ArchaeoPhases documentation built on Jan. 19, 2025, 2:33 p.m.