[MITgcm-devel] rdmnc.m or mnc ?

[Martin Losch]

Me again:
I would actually think that the two lines
>             if ismember('Zd000001' ,dims), adj = adj - 1; end
>             if ismember('Zmd000001',dims), adj = adj - 1; end
are not really necessary (they are the ones that tailor rdmnc to the  
diagnostics package). Without these lines the output of a 2d  
diagnostics field has the shape [nx ny 1 nt]. What's wrong with that?

Martin

On 18 Feb 2007, at 21:02, Martin Losch wrote:

> Hi Jean-Michel,
>
> sorry for creating this problem. I am so unfamiliar with the  
> diagnostics package, that I didn't even think about it. Please feel  
> free to revert to revision 1.10. Alternatively I can try to make  
> rdmnc work with both types of fields: [nx 1 nz nt], and [nx ny 1  
> nt]. There has got to be a way to make this work.
>
> I guess the 3rd dimension in 2d-diagnostics is necessary if you  
> write just one level of 3d field. In that case you would like to  
> know which level this is, right?
>
> Martin
>
> On 18 Feb 2007, at 01:25, Jean-Michel Campin wrote:
>
>> Hi,
>>
>> I've just added the cvs Header & Name,
>> in all utils/matlab/*.m files.
>> I don't update so often my matlab scripts, and when,
>> after an update, something is broken, it's far much easier
>> to identify which version was working and why it's broken.
>>
>> Now, regarding rdmnc.m,
>> seems to me that Revision 1.11:
>> "replace a squeeze() with a shiftdim() ..."
>> is not compatible with Revision 1.6:
>> "Fixed a bug relating to recognizing z coordinate singleton  
>> dimension"
>>
>> May be a way to fix this would be to change the mnc part of
>> diagnostics_out.F, so that 2-D diagnostics are really treated as
>> 2-D fields (without this 3rd dim Zd000001 or Zmd000001) ?
>>
>> Cheers,
>> Jean-Michel
>> _______________________________________________
>> MITgcm-devel mailing list
>> MITgcm-devel at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
>
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel