R time series object with two time dimensions

Question

When you study time series forecast you most often have to deal with two times: the time you issue your forecast and the time you make your forecast for. These two times are not regularly sampled in general.

What R object do you suggest to deal with these two times ? please indicate advantages and drawback. If no object exist, how should I use existing S4 object to create what I need?

Thanks in advance for sharing your experience in R with this type of data

Answer 1

The zoo package is useful for dealing with irregularly sampled time series.

Whether you use the issue time or the forecast time seems to be a meterological problem depend on what you're asking rather than an R problem; if you want a better answer, you'll have to provide more detail.

See the Rosetta code page on classes for a brief intro to S4. Again, better answers require more detail.

(edited to remove assumption about what you were forecasting)