Bayesthresh (version 2.0.1)

Bayesthresh: Bayesian thresholds mixed-effects models for categorical data

Description

This package fits a linear mixed model for ordinal categorical responses using Bayesian inference via Monte Carlo Markov Chains. Default is Nandran & Chen algorithm using Gaussian link function and saving just the summaries of the chains. Among the options, package allow for two other options of algorithms, for using Student's "t" link function and for saving the full chains.

Usage

Bayesthresh(formula, data, subset, na.action, A=NULL, algor = list(algorithm="NC", link="Gaussian"), Write=FALSE, priors = list(ru = 10, su = 2, dre = 20, dse = 5),  burn = 50, jump = 2, ef.iter = 4000, model=TRUE)

Arguments

formula

a two-sided linear formula object describing the fixed-effects part of the model, with the response on the left of a ~ operator and the terms, separated by + operators, on the right. The vertical bar character "|" separates an expression for a model matrix and a grouping factor.

data

an optional data frame containing the variables named in formula. By default the variables are taken from the environment from which Bayesthres is called.

subset, na.action

further model specification arguments as in lm; see there for details.

A

Matrix of variance-covariance of random effects.

algor

is a list that contains the name of the algorithm to be used. By default the algorithm is the NC with function link Gaussian

Write

the Write is a function that by default is FALSE. If TRUE, the function save the iterations of the sampling processin the matrix

priors

priors is a list that contains the parameters of the priors used to estimate the variance components of random effects

burn, jump, ef.iter

are of the arguments of iteraction. By default the burn, jump and ef.iter (effective iteractions) are 50, 2 e 4000 respectively

model

logical scalar. If FALSE the model frame in slot frame is truncated to zero rows.

Details

subset

an optional expression indicating the subset of the rows of data that should be used in the fit. This can be a logical vector, or a numeric vector indicating which observation numbers are to be included, or a character vector of the row names to be included. All observations are included by default.

na.action

a function that indicates what should happen when the data contain NAs. The default action (na.fail) prints an error message and terminate if there are any incomplete observations.

algor

the are three options algorithms, AC, MC and NC, with link function Gaussian and t-Student distribution. The object algort, by default is list(algorithm="NC", link="Gaussian").

Write

if Write=TRUE, the chain of iteractions is saved in the file output.txt. The convergence process can be analyzed by the library coda

priors

The object defines the priors for the variance components of the AC and MC algorithms. For the NC algorithm can be also defined a prior of residual variance. Objects ru (shape parameter) and su (scale parameter) are the parameters of inverse gamma for the variance components. The NC algorithm allows to change parameters of the residual variance. dre (shape parameter) and dse (scale parameter) define the prior of the residual variance. By default, algorithms AC and MC have a residual variance equal to 1.

Examples

# Not run
data(sensory)

Consumer <- factor(sensory$consumer) # Random effect
Sacarose <- factor(sensory$sacarose) # Fixed effect

#### Model 
# Not run
dex1 <- Bayesthresh(cor ~ (1|Consumer) + Sacarose, 
																				burn = 0, jump = 1, ef.iter = 10, data=sensory) 
summary(dex1)