mcmc_summary.circGLM: Obtain different central tendencies and CIs from a circGLM...

In keesmulder/CircGLMBayes: Bayesian Analysis of a Circular GLM

Obtain different central tendencies and CIs from a circGLM object

Description

Computes the mean (arithmetic or mean direction), median, and mode estimate for the MCMC chains of a circGLM object, as well as a credible interval.

Usage

mcmc_summary.circGLM(m, modebw = 0.1, ciperc = 0.95)

Arguments

m	A circGLM object.
modebw	Numeric between 0 and 1. The modes are estimated by taking the midpoint of a highest density interval. Specifically, the mode is the midpoint of the interval that contains modebw of the density of the posterior. Reasonable values are roughly between .005 and .2, although lower values may be reasonable if the number of iterations, Q, is large.
ciperc	The confidence interval percentage.

Details

The summary statistics computed have to be computed differently for linear and circular variables.

Value

A matrix with the parameters as rows, and on the columns central tendencies and appropriate credible intervals (circular quantiles and Highest Density Intervals).

Examples

dat <- generateCircGLMData()
m   <- circGLM(th ~ ., dat)
mcmc_summary.circGLM(m)

keesmulder/CircGLMBayes documentation built on July 21, 2022, 3:43 a.m.