Hi Jon et al,
IMO, this idea seems commendable for actual min/max data values.
However, is there merit in adopting somewhat more specific, albeit
longer, attribute names? I'm thinking of the situation - contrived
perhaps - wherein a non-CF-savvy end-user of a netCDF file sees a
metadata attribute called "value_min" and asks the question "Minimum
value of what?".
Perhaps the answer is self-evident from context; that is, since the
attribute is attached (in the netCDF header) to the variable it
describes then the association is clear. But in a client software
application that association might not be so obvious.
In which case is it worth considering more verbose, but more descriptive
attribute names, e.g.
{ minimum_data_value, maximum_data_value }, or
{ minimum_variable_value, maximum_variable_value }
Just a suggestion!
Regards,
Phil
On Wed, 2007-03-28 at 14:00 +0100, Jon Blower wrote:
> Dear Jonathan and list,
>
> Just to re-start this debate. The original question was whether there
> was (or could be) a pair of CF attributes to identify the min and max
> data values in a CF dataset for the purposes of aiding visualization
> tools to choose sensible colour scales.
>
<snip>
> These attributes would be optional, but easy to add to existing
> datasets, either through editing the NetCDF files or through other
> decorators such as NcML, CDML etc.
>
> I propose calling these attributes value_min and value_max.
>
> What do others think? Is this a sensible addition to CF?
>
> Jon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <
http://mailman.cgd.ucar.edu/pipermail/cf-metadata/attachments/20070328/d7a47fc9/attachment-0002.html>
Received on Wed Mar 28 2007 - 07:43:35 BST
