MarginalProba: Bayesian test for anteriority / posteriority between two...

In ArchaeoPhases: Post-Processing of the Markov Chain Simulated by 'ChronoModel', 'Oxcal' or 'BCal'

Bayesian test for anteriority / posteriority between two parameters

Description

This function estimates the posterior probability that event 'a' is older than event 'b' using the output of the MCMC algorithm. This provides a Bayesian test for checking the following assumption: "Event a is older than event b".

Usage

MarginalProba(a_chain, b_chain)

Arguments

a_chain	: Numeric vector containing the output of the MCMC algorithm for the first parameter.
b_chain	: Numeric vector containing the output of the MCMC algorithm for the second parameter.

Details

For a given output of MCMC algorithm, this function estimates the posterior probability of the event 'a' < 'b' by the relative frequency of the event "the value of event 'a' is less than the value of event 'b'" in the simulated Markov chain.

Value

An unnamed vector with the posterior probability of the assumption: "event a is older than event b"

Author(s)

Anne Philippe, [e-mail] and

Marie-Anne Vibet, [e-mail]

Examples

  data(Events); attach(Events)
  # Probability that Event.1 is older than Event.12
  MarginalProba(Event.1, Event.12)
  # Probability that Event.1 is older than Event.2
  MarginalProba(Event.1, Event.2)
  # Probability that the beginning of the phase 1 is older than the end of the phase 1
  # Should always be 1 for every phase
  data(Phases); attach(Phases)
  MarginalProba(Phase.1.alpha, Phase.1.beta)

ArchaeoPhases documentation built on June 22, 2022, 1:05 a.m.