idx: Convert indexes

In andrewzm/INLA: Functions which allow to perform full Bayesian analysis of latent Gaussian models using Integrated Nested Laplace Approximaxion

Description

Convert indexes given by to triplet '(idx, group, replicate)' to the (one-dimensional) index used in the grouped and replicated model

Usage

inla.idx(idx,
         n = max(idx),
         group = rep(1, n),
         ngroup = max(group), 
         replicate = rep(1, n),
         nrep = max(replicate)) 

Arguments

idx	The index within the basic model. (Legal values from ‘1’ to ‘n’.)
n	The length ‘n’ of the basic model.
group	The index within group. (Legal values from ‘1’ to ‘ngroup’.)
ngroup	Number of groups.
replicate	The index within replication. (Legal values from ‘1’ to ‘nrep’.)
nrep	Number of replications.

Value

inla.idx returns indexes in the range ‘1’ to ‘n*ngroup*nrep’ representing where the triplet ‘(idx,group,replicate)’ is stored internally in the full grouped and replicated model.

Author(s)

Havard Rue [e-mail]

Examples

##TODO

andrewzm/INLA documentation built on May 10, 2019, 11:12 a.m.