Geostatistical analysis under preferential samplinWe have performed, using R-inla, experiments similar to those in the Diggle et al. paper " Geostatistical analysis under preferential sampling"Here is the R code to perform a simulated example like the one in section 3 of the paper.(See ***) Here is the R code to perform a simulated example with model like the one in section 5 of the paper.(See ***) Here is the R code to re-analyse the Galicia data set (data97, data00).(See ***) (***): You may want to change the `int.strategy' and `strategy' option, to provide more accurate results for the latent field. (The current options are more appropriate in an exploratory-mode.) Also note that `inla.hyperpar()' might be required to get accurate posterior marginals for the hyperparameters, and that the `matern2d' model use zeros as boundary values hence `constr=TRUE' is required. |