Re: missing_value problem using multi-file netcdf in ferret 5.8

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hi Ingo,
It's true that the missing and fill_value flags for each variable are 
read from the first stepfile when the dataset is opened, and  applied to 
data from other stepfiles.  Are the missing /_fill_value flags different 
in your different stepfiles? 


Ingo.Bethke@student.uib.no wrote:

> Hi 
> I have a problem when using multi-file netcdf with variable scale_factor/add_offset:
>
> Only for records from the first step-file the missing_value/_FillValue flag is
> applied correctly. All other records remain unmasked. 
> Is this due to that the missing_value/_FillValue attributes are read and
> unpacked from the first stepfile only? 
> Presently, I use the first record to mask the remaining records. Is there a less
> computational expensive solution?
>
> Thanks, Ingo 
>