Skip to content
Prev 15314 / 20628 Next

How to determine the length of the required burn-in until convergence in MCMCglmm package or another package

Dear Jarrod,

So, because my zero-inflated data is a mixture of zeros and continuosly
diatributed positive values, i use two models, there are linear mixed model
(LMM) for positive values and generalized linear mixed model (GLMM) with
logit as link function for probability of positive values.
In LMM, prior for residual variance that i set is R=list(V=1, nu=0) like
you said in my post a few days ago.
And then, in GLMM, i set fix residual variance.
Is it true?

Regards

Pada tanggal 28 Mar 2017 19.20, "Jarrod Hadfield" <j.hadfield at ed.ac.uk>
menulis: