[vtk-developers] AMR warning message

[John Biddiscombe]

Burlen

OK. Thanks. The error makes a kind of sense, because the data has a 
refinenemt ratio of 1/5 per level, but the grid extent is not divisible 
by 5, so there is always a one cell overlap between low/high res parts. 
(At some point I'll try to get the people generating this data to switch 
to a better multiple)

thanks

JB

> Hi John
>
> The warning is about a condition that is typically bad for AMR codes. 
> You want coarsen and refine to be inverse operations if you operate on 
> some AMR box you want to apply the inverse and op get back what you 
> started with. When this isn't the case the coarsened grid won't cover 
> the original and In PV you'll have two level's cells visible in the 
> same location.  For the most part PV ignores the visibility arrays, 
> and I suspect that is why you don't see any changes. I am wondering if 
> your Box definition includes any ghost cells? If your image data 
> contains ghost cells, the box definitions shouldn't include the 
> ghosts. Over the next month or so we will be implementing some of the 
> missing AMR functionality as part of our co-processing SBIR. So 
> hopefully we can get these issues resolved in a timely manner.
>
> Burlen
>
>
>
> John Biddiscombe wrote:
>> I am receiving this message
>>
>> Generic Warning: In 
>> ..\..\..\..\Code\CSCS\pv-meshless\VTK\Filtering\vtkAMRBox.cxx, line 625
>>
>> This box cannot be coarsened.
>>
>>
>> when I call CreateVisibilityArrays on construction of an AMR 
>> Heirarchical box dataset. I never used to receive the warning and the 
>> code appears to still work the same.
>>
>>
>> Should I be worried?
>>
>>
>> Thanks
>>
>>
>> JB
>>
>>
>> -- 
>> John Biddiscombe,                            email:biddisco @ cscs.ch
>> http://www.cscs.ch/
>> CSCS, Swiss National Supercomputing Centre  | Tel:  +41 (91) 610.82.07
>> Via Cantonale, 6928 Manno, Switzerland      | Fax:  +41 (91) 610.82.82
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> vtk-developers mailing list
>> vtk-developers at vtk.org
>> http://www.vtk.org/mailman/listinfo/vtk-developers
>>   
>


-- 
John Biddiscombe,                            email:biddisco @ cscs.ch
http://www.cscs.ch/
CSCS, Swiss National Supercomputing Centre  | Tel:  +41 (91) 610.82.07
Via Cantonale, 6928 Manno, Switzerland      | Fax:  +41 (91) 610.82.82