Hi Mattia, > My name is Mattia Pedergnana and I am currently working at the German > Aerospace Agency in Munich. I went through the NetCDF-4 library and I would > like to kindly ask you if it possible to create different variables which > point to the same data object. If not, do you think there is any technical > issue (within the NetCDF-4 standard) by using the HDF5 library to create > them? > > Library: NetCDF-4 version 220.127.116.11 > System: Linux Suse 12.2 One of the HDF5 features omitted, to keep the netCDF enhanced data model simple, was HDF5 References. Other omitted features for which there is no netCDF-4 support are listed here: http://www.unidata.ucar.edu/software/netcdf/docs/faq.html#fv15 So if you create an HDF5 file that uses HDF5 References, you can't access them from the netCDF-4 API. You may still be able to read the data, if there is a standard variable pointing to it, but I don't know whether the netCDF-4 library just ignores other object References, or whether the presence of References in the file causes problems accessing the data by other means. If you find out by trying it, I'd be interested in the result. Although the netCDF-4 library provides no support, it may be possible to implement some of the functionality of references using attribute conventions. For example, a variable attribute named "reference_to" could name a different variable in the file, so that any scalar variable it was attached to could serve as an alias for a different variable. However, that convention would have to be supported by software that implemented the desired semantics, or by modifications to the library. The other kind of HDF5 references, pointing to a region of data elsewhere in the file, adds enough extra complexity that it would probably be better just to use HDF5, if that's what you need. It's not our intention to make netCDF-4 a complete replacement for HDF5. NetCDF-4 is just trying to fill the niche of simpler than HDF5 but not as powerful, to provide a trade-off that's better for some users ... --Russ Russ Rew UCAR Unidata Program address@hidden http://www.unidata.ucar.edu Ticket Details =================== Ticket ID: GFI-182980 Department: Support netCDF Priority: Normal Status: Closed
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.