Hi All,
Just a note to reassure everyone that I am receiving/following these recent 
comments questions and will compose a response next week after I unpack. In the 
meantime, please be aware that the ancestor search algorithm and the lateral 
search algorithm, as envisaged by the proposers, are not quite what Jonathan 
seems to think they are, so we need to do a better job of explaining them. Both 
searches should search no higher than the local apex group where the dimensions 
underlying the referring variable (and thus its coordinates) are defined. This 
is in general (and in practice) often lower down than the root group. In netCDF 
speak, the dimensions underlying the variable and the coordinate must be not 
only the same name, but the same ID (dimension IDs are unique in netCDF). 
Otherwise chaos ensues. The variable and the coordinate must both be in the 
scope (i.e., child groups) of the group where the dimension(s) are defined. 
That "feature" is due to the way that netCDF implements the Common Data Model 
(CDM). Hence ancestor search goes no higher than the group where the 
dimension(s) are defined, same as lateral search. This is the local apex group.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/cf-convention/cf-conventions/issues/144#issuecomment-429464117

Reply via email to