This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Ghansham, Can you send me a sample file that has your point data? I'd like to see what you mean by netcdf java not picking up the data variables. Thank you, Lansing > Regarding point data, pretty straight forward example would be > H.1 Point Data in CF 1.6 doc. But only change would be i will have three > different data variables (in my case u and v components of winds from 3 > different channels) along with corresponding set of lat/lon/time variables. > In current netcdf java we are not able to pick all the three set of data > variables. It picks only one. This is prett much specific to point > datasets. However in case of grids if I have 3 grids with three different > sets of lat/lon(time if present not necessary) works very well. > For second query just search GEOS satellite projection. we have support for > this in cf as far as I remember. Which particular projection classes do > this. > Ghansham > > address@hidden> wrote: > > Ghansham, > > > > If I understand your first query correctly, you have data from three > sources that perhaps nearly, but don't quite, line up in space and time. > It is certainly possible to assign dimenions lat1, lon1, lat2, lon2, lat3, > lon3 in netCDF to describe the datasets. Alternatively, if you are using > netcdf4, you can use the same names (lat and lon) for dimensions in three > different groups. I'm not sure what you mean when you say that netcdf-java > only picks up one of the datasets. > > > > The best answer on how to package your data is to suggest looking at it > from the perspective of the end user or application, as ultimately, this is > what matters most. For instance, you could fudge the three dataset grids > into an average and report only one lat/lon grid, or you could trust the > end user to interpret the three grids appropriately. > > > > As to your second query, can you clarify what you need or are trying to > do? > > > > Regards, > > Lansing Madry > > Unidata > > Boulder, Colorado > > > > > > > >> Sir, > >> > >> > >> First query is related to point data... > >> > >> I have winds from three channels (VIS/IR/WV) which have three different > >> sets of lat/lon. I am trying to follow CF conventions. Is it possible > >> to pack all the three sets of winds in one hdf/netcdf file and then read > >> them using netcdf-java enhanced data model... What I am seeing is that > >> it picks only one of the datasets? Or is it possible using CDM.... > >> > >> Second query is related to GEOS Projection. > >> > >> Is there any class in ucar.unidata.geoloc.projection and corresponding > >> class in ucar.nc2.dataset.transform. I think it is already in there in > >> the CF conventions. > >> > >> I can try writing if it is not available. > >> > >> regards > >> Ghansham > >> > >> > > > > > > Ticket Details > > =================== > > Ticket ID: KWA-143788 > > Department: Support netCDF Java > > Priority: Critical > > Status: Open > > > > > > Ticket Details =================== Ticket ID: KWA-143788 Department: Support netCDF Java Priority: Critical Status: Open