Introducing a CF domain variable#302
Conversation
ch05.adoc
Outdated
| order in which the dimensions appear in the **`dimensions`** attribute | ||
| string. The presence of a **`dimensions`** attribute will identify the | ||
| variable as a domain variable. If a domain has no dimensions then the | ||
| value of the **`dimensions`** attribute must be an empty string. |
There was a problem hiding this comment.
Can a domain exist without any dimensions?
There was a problem hiding this comment.
Yes, it can - just provide the dimensions with a an empty string (domain:dimensions = "" ;). It still needs the dimensions attribute so that it is recognisable as a domain variable.
Such a domain can still have scalar coordinate variables, grid mappings, etc. I mentioned that it is allowed in the conformance requirements, but I agree that it should be mentioned explicitly in the new 5.8 text. I'll propose some new words in this thread, shortly ...
There was a problem hiding this comment.
(I should have read the existing text first rather than replying from memory - sorry I restated a lot of uncontested stuff. I'll try again)
A domain can certainly exist without explicit dimensions, as they may be implied by scalar coordinate variables.
char domain ;
domain:dimensions = "" ;
domain:cooridnates = "time" ;
double time ;
time:units = "days since 2000-01-01" ;
Would this next one be a valid domain, though?
char domain ;
domain:dimensions = "" ;
domain:grid_mapping = "lambert_conformal" ;
My gut feeling is "yes" ....
|
If no further discussion takes place in #301, we will merge this pull request on Monday November 23rd. |
JonathanGregory
left a comment
JonathanGregory
left a comment
There was a problem hiding this comment.
I think the content is fine, thanks, but I would note two minor problems in wording:
-
"blank separated list of the dimensions names" should be dimension or dimensions'
-
In "Multiple domain variables may exist in a file with, or without, other data variables" I think "other" should be deleted, because domain variables are not data variables.
Jonathan
|
@davidhassell -- please update the history then we can go ahead and merge IMHO. |
|
Thanks, @dblodgett-usgs. History is updated. |
4 participants
See issue #301 for discussion of these changes.
Release checklist
cf-conventions.adoc?cf-conventions.adocup to date? Versioning inspired by SemVer.history.adocup to date?For maintainers
After the merge remember to delete the source branch.
Tags are set at the conclusion of the annual meeting; until then
masteralways is a draft for the next version.