CARBayes (version 5.2.5)

summarise.lincomb: Compute the posterior distribution for a linear combination of the covariates from the linear predictor.

Description

This function takes in a CARBayes model object and computes the posterior distribution and posterior quantiles of a linear combination of the covariates from the linear predictor. For example, if a quadratic effect of a covariate on the response was specified, then this function allows you to compute the posterior distribution of the quadratic relationship.

Usage

summarise.lincomb(model, columns=NULL, quantiles=0.5, distribution=FALSE)

Arguments

model

A CARBayes model object from fitting one of the models in this package.

columns

A vector of column numbers stating which columns in the design matrix of covariates the posterior distribution should be computed for.

quantiles

The vector of posterior quantiles required.

distribution

A logical value stating whether the entire posterior distribution should be returned or just the specified quantiles.

Value

quantiles

A 2 dimensional array containing the requied posterior quantiles. Each row relates to a data value, and each column to a different requested quantile.

posterior

A 2 dimensional array containing the requied posterior distribution. Each column relates to a different data value.

Examples

# NOT RUN {
## See the vignette accompanying this package for an example of its use.
# }