[MITgcm-devel] diagstats_mnc_out

[Jean-Michel Campin]

Hi,

I think this is related to the "problem" I mentionned in the
previous reply (writing all of them and not the selected ones),
and when writing output to ASCII file, you only get the regions you 
selected, see the test exp. aim.5l_cs.thSI, the 2 output files:
landStDiag.0000000000.txt
thSIceStDiag.0000000000.txt

Jean-Michel

On Wed, Jul 30, 2008 at 05:48:56PM +0200, Martin Losch wrote:
> well, 'user error' ...
> It's kind of unintuitive that "stat_region(1,1)" in data.diagnostics  
> always requires the entry "0" for the global mean to give reasonable  
> values and not produce some garbage (1e86 or 1e-317 or so) ...
>
> Martin
>
> On 30 Jul 2008, at 17:40, Martin Losch wrote:
>
>> sorry, user error.
>> M.
>> On 30 Jul 2008, at 16:52, Martin Losch wrote:
>>
>>> Hi there,
>>> I have a problem with diagstats output for mnc with regional  
>>> statistics: it looks like the global value (index 0)  is not  
>>> properly written by the mnc-pkg for any of my diagstats-variables.  
>>> The ascii output is OK. I cannot see an obvious reason in  
>>> diagstats_mnc_out.F
>>>
>>> Before I dig into this and try it on various platforms: Has anyone  
>>> encountered this problem before?
>>>
>>> Martin
>>>
>>> _______________________________________________
>>> 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
>
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel