Hello Karl,
An astute point. I confess to sweeping this under the carpet a little
bit and ignoring (deleting) the standard CF global attributes as given
in appendix A, but insisting that non-standard ones are identical for
aggregated fields.
This is an easy and pragmatic approach as a programmer, but probably
not the most useful one? For me, it raises issues such as whether or
not an attribute should be a scalar auxiliary coordinate, or could
perhaps be 'promoted' to be one during the aggregation. Or perhaps
such attributes should be concatenated? etc. This area is
deliberately not regulated in our CF aggregation rules - rather
leaving it to an implementation's, or the users, discretion, which can
vary with application.
Since a tool should be useful to the users (not just easy for
programmers!), opinions are welcome.
All the best,
David
---- Original message from Karl Taylor (10AM 19 Mar 12)
> Date: Mon, 19 Mar 2012 10:51:36 -0700
> From: Karl Taylor <taylor13 at llnl.gov>
> User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.5; rv:10.0.2)
> Gecko/20120216 Thunderbird/10.0.2
> To: cf-metadata at cgd.ucar.edu
> Subject: Re: [CF-metadata] cf2cf - a new tool
>
> presumably this tool is limited to combining files with identical
> global attributes? Of course the global attributes should conflict
> between variables that you want to aggregate.
>
> Karl
>
> On 3/19/12 10:28 AM, David Hassell wrote:
> >Hello,
> >
> >I'd like to advertise a hopefully useful new tool called 'cf2cf' for
> >combining many CF-netCDF files into a single CF-netCDF file, where the
> >input fields are aggregated into as few multidimensional netCDF
> >variables as possible.
> >
> >This is a command line utility which utilizes the proposed CF
> >aggregation rules (trac ticket #78) to combine CF fields based on
> >their metadata. In particular differing variables' units, dimension
> >order, dimension direction and netCDF variable names are are accounted
> >for in the aggregation process.
> >
> >A quick, simple example:
> >
> >% cf2cf -o one_output_file.nc lots_of_input_files*.nc
> >
> >All details, including full documentation, more examples and
> >installation, may be found at
> >http://www.met.reading.ac.uk/~david/cf2cf.html
> >
> >This tool uses the python cf package
> >(http://code.google.com/p/cf-python), which also makes the same CF
> >aggregation and the CF data model (trac ticket #68) available in an
> >API, in a memory efficient manner.
> >
> >Any feedback, positive or negative, on cf2cf, the CF aggregation rules
> >and the cf python package would, of course, be most gratefully
> >received, to help make these better.
> >
> >All the best,
> >
> >David
> >
> >--
> >David Hassell
> >National Centre for Atmospheric Science (NCAS)
> >Department of Meteorology, University of Reading,
> >Earley Gate, PO Box 243,
> >Reading RG6 6BB, U.K.
> >
> >Tel : 0118 3785613
> >Fax : 0118 3788316
> >E-mail: d.c.hassell at reading.ac.uk
> >_______________________________________________
> >CF-metadata mailing list
> >CF-metadata at cgd.ucar.edu
> >http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
--
David Hassell
National Centre for Atmospheric Science (NCAS)
Department of Meteorology, University of Reading,
Earley Gate, PO Box 243,
Reading RG6 6BB, U.K.
Tel : 0118 3785613
Fax : 0118 3788316
E-mail: d.c.hassell at reading.ac.uk
Received on Mon Mar 19 2012 - 14:16:36 GMT
