Due to the current gap in continued funding from the U.S. National Science Foundation (NSF), the NSF Unidata Program Center has temporarily paused most operations. See NSF Unidata Pause in Most Operations for details.
Eugene, >Date: Mon, 31 Jan 2000 10:19:09 +1100 (AEDT) >From: Eugene Tan <chet@xxxxxxxxxx> >Organization: . >To: visad-list@xxxxxxxxxxxxx >Subject: Re: Save problem? >Keywords: 200001302324.QAA05731 In the above message, you wrote: > I tried to save my FieldImpl to a netcdf file but it keeps crashing out > with the following exception: > > java.lang.ClassCastException: visad.Irregular1DSet > at > visad.data.netcdf.out.VisADAdapter.define1DDim(VisADAdapter.java:388) > at > visad.data.netcdf.out.VisADAdapter.defineSampledSetDim(VisADAdapter.java:361) > at visad.data.netcdf.out.VisADAdapter.visit(VisADAdapter.java:269) > at visad.data.netcdf.out.VisADAdapter.visit(VisADAdapter.java:112) > at visad.data.netcdf.out.VisADAdapter.<init>(VisADAdapter.java:68) > at visad.data.netcdf.Plain.save(Plain.java:86) > at newTest.main(newTest.java:42) > > The code looks like this: > > Plain plain = new Plain(); > plain.save("toGraph.nc", field, true); > > field is a FieldImpl. > > I found out that it works if I don't resample the field first. The saving > stops working after I resample my field. (I use an Irregular1DSet in the > resampling). The netCDF export module can only save a subset of all possible VisAD data objects. One of the objects it can't save is a FieldImpl with a Irregular1DSet domain. The reason for this is because the coordinate variables in a netCDF dataset should be sorted. I'm going to modify the code so that Plain.save(String) throws a BadFormException in this case. In any case, you should modify your code so that the FieldImpl has a Gridded1DSet for the domain. Regards, Steve Emmerson <http://www.unidata.ucar.edu>
visad
archives: