[ncl-talk] Suppress "attempt to fix non-monotonic aggregation variable"

[Dave Allured - NOAA Affiliate]

Joe, that error comes out of the automatic aggregation feature of addfiles
"cat" mode (concatenate).  Addfiles also has a "join" mode which handles
dimensions differently, and does not do coordinate aggregation.  Read about
that in function docs for addfiles and ListSetType.  There are examples in
both function docs.  Will join mode work for your application?


On Sun, Oct 23, 2022 at 1:01 PM Dennis Shea via ncl-talk <
ncl-talk at mailman.ucar.edu> wrote:

> x("time/record",[level, latitude,longitude)
>
> I speculate that NCL's developers were trying to prevent a user from
> 'shooting themselves in  the foot'
> I an not aware any option to disable this 'feature' of NCL
>
> ==================
>
> ***Possibly:*** Non-NCL
>
> *Climate Data Operators* <https://code.mpimet.mpg.de/projects/cdo>
>
> *https://www.unidata.ucar.edu/software/netcdf/workshops/2012/third_party/CDO.html*
> <https://www.unidata.ucar.edu/software/netcdf/workshops/2012/third_party/CDO.html>
>
> %>  cdo cat *.nc out.nc
>
> ==========
> *netCDF Operators* <https://nco.sourceforge.net/>
> see *ncrcat *examples
>
> On Sat, Oct 22, 2022 at 3:26 PM Joseph Clark via ncl-talk <
> ncl-talk at mailman.ucar.edu> wrote:
>
>> Hi NCL users,
>>
>> I am using NCL version 6.3.0.  I have a line like,
>>
>> *variable = ncfile[:]->varname*
>>
>> I get the following error,
>>
>>
>>
>> *warning:error attempting to fix non-monotonic aggregation variable*
>> I understand exactly what the error is saying, but I do not want the
>> concatenation to change the ordering of the leftmost dimension. It is the
>> correct order. I know I could make a change to the netcdf files to have a
>> time array and time unit that NCL understands is proper order.
>>
>> * However, it would be simplest if NCL did not "attempt to fix
>> non-monotonic aggregation variable" opting instead to ignore what the order
>> of the time array is. Is there a way to tell NCL to ignore the
>> non-monotonicity? *
>> Thanks!
>>
>> Joe
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.ucar.edu/pipermail/ncl-talk/attachments/20221023/c6988ccb/attachment.htm>