[Dune] parallel coarsening

[Marco Cisternino]

Hi Andreas,
I wrote a small code which replicates the weird adapting process I have 
in my main code.
As attachment you have an archive containing the main 
(adaptFromRestore.cc) and 3 input files in the folder data_1/ (2 for the 
grid and one for the data).
The code is quite simple:
it restores my grid with marks attached as user data (0-field is for 
marks, 1-field is for the indices, 2-field is for the rank),
it plots the data in a vtk file in order to verify y-symmetry and 
z-uniformity of the marks and to see the starting grid
it marks the elements using the 0-field in the data structure, where the 
marks have been stored
it adapts and plots the grid.

When I wrote you last time, I thought it was a matter of parallel 
implementation, but it is not.
I have exactly the same behaviour with 1 process too.
That' s why I'm sending you input data for 1 process run. I have input 
data for 8 procs, if you want them, just ask.

I use dune 2.3 and alugrid 1.52. The run is fast and the grid is quite 
light.
As you will see the marks are symmetric in y and uniform in z on the 
staring grid, but the adaptation is not.
Could this be a bug in alugrid?? I hope it's not.
Thanks a lot for your help.

Bests,
Marco



-- 
------------------------------------
Marco Cisternino, Ph.D.
OPTIMAD Engineering s.r.l.
Via Giacinto Collegno 18
10143 Torino - Italy
www.optimad.it
marco.cisternino at optimad.it
+39 011 19719782
------------------------------------

-------------- next part --------------
A non-text attachment was scrubbed...
Name: weirdAdaptationSmallCode.tar.gz
Type: application/x-gzip
Size: 285332 bytes
Desc: not available
URL: <https://lists.dune-project.org/pipermail/dune/attachments/20130612/05e3e324/attachment.bin>