Re: CMOR

Hi Mark,

I'm not sure if anybody responded to this.

This is the first request that I've seen for the CMOR functions to be wrapped in NCL.

Are you familiar with WRAPIT, which allows you to "wrap" Fortran subroutines so you can call them
from NCL?

There are some simple examples and full documentation at:

http://www.ncl.ucar.edu/Document/Tools/WRAPIT.shtml

I looked at these functions Friday, and I seem to remember there is only a Fortran 90 interface?

If so, you'll need to create little Fortran 77 stub files, because WRAPIT only works on Fortran 77.
These are pretty simple to do if you know Fortran 77 and Fortran 90.

See example 1 on this page, which shows how to use a stub file (although it doesn't need to).

I'll create a ticket for your request, in case it comes up again.

--Mary

On May 27, 2010, at 10:37 PM, mark collier wrote:

> Hi,
> I know that there is no public NCL interface (bindings) to CMOR (http://www2-pcmdi.llnl.gov/cmor
> ) but wondering if someone has contemplated doing this given that for
> many of us NCL could offer a very convenient scripting language for
> processing raw model data, whereas CMOR handles mostly the metadata
> requirements of the output? I'm afraid my ability to write external
> functions in NCL is non existent. Alternatively it might be just
> easier to write a pure NCL system that handles everything, including
> looking up the CMIP5 tables? We've struggled with the python CMOR
> interface although have had some limited success.
> Regards,
> Mark.
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Tue Jun 1 09:10:39 2010