IMO, missing_value was likely incorrectly deprecated in the netcdf user guide, and CF incorrectly
echoed that decision. Russ and I are reviewing that, and if we decide thats the case, we will
correct the users guide and perhaps make a recommendation to CF.
Yes, nj library will recognize multiple missing_values.
BTW, with the advent of version 4, im starting to use "nj (and/or) cdm library" instead of "nj22".
bad idea to use version numbers, since versions change!
Jon Blower wrote:
missing_value is deprecated by CF in favour of _FillValue but I guess
this is a legitimate use. Will nj22 recognise these multiple values
for missing_value?
Jon
On Wed, May 28, 2008 at 8:44 PM, John Caron <caron@xxxxxxxxxxxxxxxx> wrote:
The "missing_value" attribute allows multiple values. I prefer it over
_FillValue myself, but it may not be as widely recognized by apps.
Jon Blower wrote:
Hi all,
Oops, sorry, I misunderstood something - looking at the CF document it
seems that only one FillValue is allowed (not an array of values).
This seems to be an issue with the source data, not with netcdf-java
or CF.
Cheers, Jon
On Wed, May 28, 2008 at 10:41 AM, Jon Blower <jdb@xxxxxxxxxxxxxxxxxxxx>
wrote:
Hi all,
Sorry if this question has been asked before, but we have a situation
where we have source NetCDF files that have two missing values (-1 and
-999). Only one of these seems to be recognized by Java-NetCDF
(2.2.22) (using VariableEnhanced.convertScaleOffsetMissing(val)). Is
there a way to have both of these values accepted as missing values?
Mike Grant (cc-ed) can provide example data or CDL if required.
Cheers, Jon
--
--------------------------------------------------------------
Dr Jon Blower Tel: +44 118 378 5213 (direct line)
Technical Director Tel: +44 118 378 8741 (ESSC)
Reading e-Science Centre Fax: +44 118 378 6413
ESSC Email: jdb@xxxxxxxxxxxxxxxxxxxx
University of Reading
3 Earley Gate
Reading RG6 6AL, UK
--------------------------------------------------------------
