[R-sig-Geo] ZIP or Hurdle model in BRugs

Karen Updegraff kupdegra at d.umn.edu
Thu Oct 4 16:11:27 CEST 2012


I have some zero-inflated count data that I would like to use for
small-area spatial interpolation (of plant population densities).  As
GeoR and GeoRglm don't appear to accommodate ZIP or hurdle models, I'm
thinking the R-OpenBugs binding (BRugs) might be a plausible route for
setting up and evaluating such a model. However, as a complete
OpenBUGS neophyte I am a bit overwhelmed by the coding requirements.
None of the examples provided with OpenBUGS or GeoBUGS have been
illuminating.   One problem is that the OpenBUGS documentation for
linux (cmd line system) is nonexistent.  I understand that JAGS is the
recommended alternative but don't believe it has a spatial module.

If anyone has some example code that approximates what I am trying to
do (fit a model to spatially-referenced count data, then run the krige
in GeoBUGS?) and would be willing to share it, it would be a huge
help.

thanks, Karen



More information about the R-sig-Geo mailing list