[ncl-talk] ncclose: ncid 131072: NetCDF: Can't open HDF5 attribute

Mary Haley haley at ucar.edu
Fri Aug 24 12:16:40 MDT 2018


Ehsan,

Can you provide the original
era5precjan2018t6s1_12.nc file?
If it's not too large, but too large for email, then you can use ftp:

http://www.ncl.ucar.edu/ftp_files.shtml

Thanks,

--Mary


On Wed, Aug 22, 2018 at 11:30 PM, Ehsan Taghizadeh <
ehsantaghizadeh at yahoo.com> wrote:

> Dear Dennis
> Thank you so much for your nice reply.
> As you said I've encountered abnormally size before in ncl-6.4. However in
> ncl6.5 I encountered addition problem and in this email I mean this new
> problem (may bug in ncl6.5). It seems ncl6.5 reset wrong date/time. For
> example I have "0103-01-01 00:00:00" instead of "2018-01-01 07:00:00" for
> "Date     Time".
>
> Sincerely
> Ehsan
> On Thursday, August 23, 2018, 2:58:34 AM GMT+4:30, Dennis Shea <
> shea at ucar.edu> wrote:
>
>
> This [netCDF-4 file size] has been discussed in ncl-talk before. Please
> search the archives.
> --
> It has to do with the netCDF-4 '*chunk size*' used by NCL.  A JIRA ticket
> has been opened.
>
> NCAR Command Language / *Defect NCL-2757*
> *NetCDF-4 file with unlimited time dimension size has abnormally large
> size *
>
> You can change the chunk size: see
>
> *http://www.ncl.ucar.edu/Docume nt/Functions/Built-in/filechun
> kdimdef.shtml*
> <http://www.ncl.ucar.edu/Document/Functions/Built-in/filechunkdimdef.shtml>
>
> You may have to experimet.
> ======
>
> *I suggest trying just writing netCDF-3 [the default]*
> Good luck
>
>
> On Tue, Aug 21, 2018 at 8:37 AM, Ehsan Taghizadeh <
> ehsantaghizadeh at yahoo.com> wrote:
>
> Hi,
> I used attached script (prc_sum_write.ncl) which has been provided by
> ncl-talk to sum precipitation in different time steps. In ncl-6.5 I
> encountered following message:
>     ncclose: ncid 131072: NetCDF: Can't open HDF5 attribute
> It seems the script does summation correctly, however reset wrong
> date/time! (cdoout6.5.dat in attached files)
> The script works correctly in ncl-6.4; however the problem with ncl-6.4
> was about following line:
>     filedimdef(fo,"time",-1,True)      ; make time UNLIMITED
> which produced a much bigger size output. Same is true in ncl-6.5 which
> the output without above line will have 755 KB size and the output with
> above line will have 24.0 MB. However the date/time in output file for
> ncl-6.4 are correct (cdoout6.4.dat in attached files). I also attached
> output files from both version of ncl in small size (commented
> filedimdef(fo,"time",-1,True)) .
> Should I use ncl-6.4 with this script or I did something wrong? I'll be
> thankful for any help.
>
>
> Sincerely
> Ehsan
>
> ______________________________ _________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/ mailman/listinfo/ncl-talk
> <http://mailman.ucar.edu/mailman/listinfo/ncl-talk>
>
>
>
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20180824/4b906a12/attachment.html>


More information about the ncl-talk mailing list