There is no "delete" function (yet, maybe one will appear later), but you can just copy all the stuff you want to keep to a new file.
The basic workflow is: get all the "object_ids" from the file, read all these quantities, figure out which to keep and not, and write those you want to keep to a new NC file. Simple code (hope you get the idea):
old_filename = 'analysis.nc'
new_filename = 'analysis2.nc'
delete_these = ['gID001']
#-------------------------------------------------
from NL.IO.NLSaveUtilities import nlinspect
objs = nlinspect(old_filename)
all_ids = [i[1] for i in objs]
for id in all_ids:
print id,
if id in delete_these:
print 'skippping'
else:
print 'including'
obj = nlread(old_filename,object_id=id)[0]
nlsave(new_filename,obj,object_id=id)
Inspecting the object_ids to find out which to keep or remove is most easily done in VNL.
Note: the script above uses the undocumented function
nlinspect; you can not trust its interface to remain constant in future editions of ATK, but for now it works this way.
Play around with it!
