LN_hier_existence {BayesLN}

Numerical evaluation of the log-normal conditioned means posterior moments

Description

Function that evaluates the existence conditions for moments of useful quantities in the original data scale when a log-normal linear mixed model is estimated.

Usage

LN_hier_existence(X, Z, Xtilde, order_moment = 2, s = 1, m = NULL)

Arguments

X

Design matrix for fixed effects.

Z

Design matrix for random effects.

Xtilde

Covariate patterns used for the leverage computation.

order_moment

Order of the posterior moments required to be finite.

s

Number of variances of the random effects.

m

Vector of size s (if s>1) that indicates the dimensions of the random effect vectors.

Details

This function computes the existence conditions for the moments up to order fixed by order_moment of the log-normal linear mixed model specified by the design matrices X and Z. It considers the prediction based on multiple covariate patterns stored in the rows of the Xtilde matrix.

Value

Both the values of the factors determining the existence condition and the values of the gamma parameters for the different variance components are provided.