Skip to content
Prev 3195 / 20628 Next

Boosting computation time of glmmPQL when specifying spatial, correlation structure

On 17/01/2010 11:00, r-sig-mixed-models-request at r-project.org wrote:
You are not splitting up the data set. Instead you are imposing the 
correlation on a sub-block of data.
The first thing to do is to use decent starting values for the range 
(and nugget???). See
?corSpher

Alain