Forms the expected covariance structure of link-scale observations for GLMMs fitted with MCMCglmm
buildV(object, marginal=object$Random$formula, diag=TRUE, it=NULL, posterior="mean", ...)If diag=TRUE an n by n covariance matrix. If diag=FALSE and posterior!="all" an 1 by n matrix of variances. If posterior=="all" an nit by n matrix of variances (where nit is the number of saved MCMC iterations).
an object of class "MCMCglmm"
formula defining random effects to be maginalised
logical; if TRUE the covariances betwween observations are not calculated
integer; optional, MCMC iteration on which covariance structure should be based
character; if it is NULL should the covariance structure be based on the marginal posterior means ('mean') of the VCV parameters, or the posterior modes ('mode'), or a random draw from the posterior with replacement ('distribution'). If posterior=="all" the posterior distribution of observation variances is returned
Further arguments to be passed
Jarrod Hadfield j.hadfield@ed.ac.uk
MCMCglmm