fitMBH: fitMBH

In susanjarvis501/MBH: Model-based hypervolumes

Description

This function fits modelled based or empirical hypervolumes to multivariate data

Usage

fitMBH(x, vars = c("V1", "V2", "V3"), groups = "Group", nc = 3,
  ni = 1e+05, nb = 20000, nt = 20)

Arguments

x	Data to fit hypervolume to
vars	Names of variables in x to use for hypervolume construction
groups	Name of the grouping variable in x. Use NULL if no groups present or to ignore grouping structure and fit an empirical hypervolume
nc	Number of MCMC chains
ni	Number of MCMC iterations (default 100000)
nb	Length of burnin
nt	Thinning paramter

Details

To use the coda package for mcmc diagnostics, you first need to convert the samples to mcmc.list format. This can be completed with the as.mcmc.list function from the mcmcr package. To inspect the mcmc chains for the estimated covariance matrix use plot(as.mcmc.list(m3$samples$tau))

Value

means - Estimated means of each variable

covariance - Estimated covariance structure

volume - Estimated hypervolume size

group_means - Estimated group means

group_variances - Estimated between-group variances for each variable

samples - The output from the jags.samples function

susanjarvis501/MBH documentation built on Aug. 27, 2020, 7:37 a.m.