NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Dear Simone,if I understand you correctly you want to use CDO to merge files with different variables order. I had the same problem as i was merging data coming from ECMWF weather forecasts, where the data are not always in the same order.. ( due to the differnet run time of the program codes) I talk to uwe, and the solution we came up is: use selcode to write out a variable in a loop and add back to a final file, which then contains all variables in order. ( go over all files..
If you need a csh/cdo script for that, i already have one around. hannes At 07:34 PM 5/29/2007, Simone Marras wrote:
Thank you Philippe for the answer and the hint. I tried all the options i have, and I am starting to think that it is not the time that is affecting the CDO merge option, but some other variables in my GISS files, because even by changing the date position by running the *.c generated code (coming from "ncgen file > file.c") I still obatin a merged file that has all the variables, but the time is not there anymore. In case anyone else had the same problem using CDO mergetime, please, let me know! Regards and thanks simone On Tuesday 29 May 2007 17:01, Philippe Poilbarbe wrote: > Simone Marras a écrit : > > ... > > Is there a way to control the position inside the file where the variable > > value is added? > > I need to put it in a certain position because when I use CDO to merge > > several files, the output merged file does not have a date anymore. > > ... > > No there is no way to control the position: they are in > the order they have been created (with a little exception > for 'record' vars). > Anyway no program should take into account the position > of the variable (in fact linked to the var Id) but only > the names. > > For CDO (I don't know it except its name), it is a bit > strange that a "generic" program throw away a variable > due to its position. > It may be related to some CDO options usage or the way > it is working (I did have some problems like this > with the NCO programs): check all options. > > To see if it is related to the position, you can try > to export your NetCDF file with ncdump, then with a > text editor move the variable (in the header only, it > is sufficient) where your think it should be and then > regenerate the file with ncgen. Rerun CDO with this > new file and it is probable that you will have the > same result (and normally it should be the same). > > If it is not the same result (except, maybe for > variable positions if the resulting file), it may > be a CDO bug. -- Simone Marras, Ph.D. Student BSC & Universitat Politecnica de Catalunya Jordi Jirona 29 08034 Barcelona Spain ============================================================================== To unsubscribe netcdfgroup, visit: http://www.unidata.ucar.edu/mailing-list-delete-form.html ==============================================================================
Hannes Isaak ReuterInstitute for Environment and Sustainability - Land Management and Natural Hazards Unit - European Commission - DG Joint Research Centre
TP 280, Via Fermi 1, I - 21020 Ispra (VA), Italytel +39-0332-78.5535, fax +39-0332-78.6394, Cell:+39-3479720381, e-mail: hannes.reuter@xxxxxx
www: http://eusoils.jrc.it/"The views expressed are purely those of the writer and may not in any circumstances be regarded as stating an official position of the European Commission."
============================================================================== To unsubscribe netcdfgroup, visit: http://www.unidata.ucar.edu/mailing-list-delete-form.html ==============================================================================
netcdfgroup archives: