#6 Improve Metadata

[Adam M. Wilson]

It would be nice if the mcmc functions included the fitting metadata in the returned object to make it easier to archive model results via the hSDM.nc function. Currently you have to feed hSDM.nc the model formula and other metadata again (which increases chance of a coding error). It would be nice to have the terms used in the model fitting attached to the returned object (in separate slots or just in the list of results):
1) model formulas (suitability and observability)
2) model type (ZIB, ZIB.icar, etc.)
3) model settings (prior values, etc)
4) mcmc settings (burnin, mcmc, thin)
5) model run time (perhaps?)
6) also, what if we could add a metadata list to the fitting procedure (species name, group, model name, etc.). When fitting multiple species and multiple models it would be nice to have some way to carry the metadata along the chain from model fitting -> model posteriors -> validation/performance metrics -> predictions. If we could add some optional metadata list to the original fitting call (e.g. hSDM.ZIB), it would be easier to keep track of model results archived in the netcdf output.

We should probably follow how these things are attached to lm() objects...