On Thu, 3 Apr 2014, Roy Mendelssohn wrote:
The state-space approach has the advantage in the appropriate situations that you can model the trends and seasonals and cycles in a way that doesn't assume stationarity and provides a lot of flexibility. To me a lot of it depends on if the nature of the irregularity is an inherent property of the data themselves or of the observation process - for example if it makes sense to say the physical observable is there every month but we just have not been able to observe. I have fit state-space models to reasonably sparse data with what appear to be good results.
Roy, Hadn't thought about the nature of the irregularity. For all my data, the nature is observational. Water and air chemistry (and physical parameters such as flow and temperature) are always there. Biota might be different. Plants tend to not move too quickly so they're always present ... or not. But, animals (aquatic and terrestrial) may not be present or may be present but not observed.
If you want I can send you some examples off-line.
That would be much appreciated. Thanks, Rich ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
