Rich Lysakowski writes:
> I wish we had a mechanism beyond conventions for grouping multiple
> netCDF file objects into more complex objects that could be mailed/copied
> around the network (a directory type structure would be fine for now).
> Tarring files together doesn't let you access what's in the files easily.
> Does anyone have a better approach now?
Our Envision software project addresses multiple file problems for a
data set by defining "projects" which describe variables and
dimensions that are divided over multiple files. It works by letting
the user specify which variables and dimensions correspond in the
various files. Once defined in a project, these "logical" objects can
then be accessed as if they were in a single file.
Other dimensions may be implied by the division of the data into
files, such as when each file represents one time measurement. In
this case, the time dimension may be defined only by a unique file
name, e.g. 28Oct1400Z.cdf. Our logic allows a "virtual" dimension to
be defined like other logical dimensions. This allows the user to
select time ranges over multiple files, even when no time dimension is
stored with the data.
The multiple file issue is really just higher-level metadata
management and it requires only one project file to store this new
metadata. The netcdf files themselves are unaffected. With this
approach, users are free to use any conventions they please, but can
still manage a segmented data set in a more convenient way. Our code
is still being developed, but will be freely available early next
year.
--------------------------------------------------------------------------
Keith Searight, Research Programmer keith@xxxxxxxxxxxxxxxxxxxx
Univ. of Illinois, Atmospheric Sciences Ph. (217) 333-8132
105 S. Gregory Ave., Urbana, IL 61801 Fax (217) 244-4393
--------------------------------------------------------------------------
