[CF-metadata] Comment in coordinate variables (and CF checker)

From: r.s.hatcher at reading.ac.uk <r.s.hatcher>
Date: 23 Apr 2012 15:51:11 +0100

Hi Nan,

The CF checker is producing the INFO message because according to Appendix
A the "comment" attribute can be attached to either a (G)lobal variable or
a (D)ata variable containing non-coordinate data.

This is just an informational message alerting you to the fact that you may
be using the attribute incorrectly. As far as I am aware there is nothing
to stop you using the comment attribute on a coordinate variable (hence the
INFO message not WARNING or ERROR) - it just isn't a standard use of it.

Regards,
Ros.


On Apr 23 2012, Nan Galbraith wrote:

>Hi all -
>
>I've got a question about the correct use of comments in CF.
>
>In chapter 2 of the spec (actually, 2.6.2. Description of file
>contents) the term comment is defined this way: "Miscellaneous
>information about the data or methods used to produce it."
>
>I'm using comments to describe some of my coordinate variables:
> DEPTH:comment = "approximate instrument depth" ;
> LATITUDE:comment = "surveyed anchor position" ;
> LONGITUDE:comment = "surveyed anchor position" ;
>
>This SEEMS like what this attribute is meant to do; it gives an idea
>of the coordinates' accuracy, or lack thereof, which can't really
>be quantified.
>
>These produce a warning from the NCAS-CMS (NERC) CF checker:
>' INFO: attribute 'comment' is being used in a non-standard way.'
>
>Isn't this the correct use of the comment field? Is there some prohibition
>against this, or is there another use of the comment attribute, that
>restricts its 'normal' use? I've pored over the spec, and can't figure
>out why this could be considered 'non-standard'. Any ideas? Am
>I missing something?
>
>Thanks -
>
>Nan
>
>
Received on Mon Apr 23 2012 - 08:51:11 BST