DOLFIN

HarmonicSmoothing::move is broken

Bug #1047641 reported by Garth Wells
16
This bug affects 3 people
Affects Status Importance Assigned to Milestone
DOLFIN
Fix Committed
Medium
Jan Blechta

Bug Description

The function HarmonicSmoothing::move is broken because somewhere makes an assumption on the dof ordering.

See original description

Related branches

lp:~blechta/dolfin/harmonic_smoothing
Anders Logg: Pending requested
Diff: 68 lines (+4/-13)
4 files modified
demo/undocumented/smoothing/python/demo_smoothing.py (+1/-4)
demo/undocumented/submesh/cpp/main.cpp (+0/-3)
demo/undocumented/submesh/python/demo_submesh.py (+1/-4)
test/regression/test.py (+2/-2)
Garth Wells (garth-wells)
Changed in dolfin:
assignee: nobody → Anders Logg (logg)
Garth Wells (garth-wells)
description: updated
Revision history for this message
Yaakoub El Khamra (yelkhamra) wrote :

With this bug unresolved I cannot use my 2D domain with a hole in it code. It does a snap_boundary same as in the smoothing example. Should I just go with mesh generation, an external mesh generator or are there alternatives? Any suggestions?

Thanks in advance
Regards
Yaakoub

Revision history for this message
Anders Logg (logg) wrote :

I have made a temporary fix so that an error is only thrown when the parameter "reorder_dofs_serial" is set to True. If set to False, the function works fine.

Changed in dolfin:
status: New → Confirmed
importance: Undecided → Critical
importance: Critical → Medium
Jan Blechta (blechta)
Changed in dolfin:
assignee: Anders Logg (logg) → Jan Blechta (blechta)
status: Confirmed → In Progress
Jan Blechta (blechta)
Changed in dolfin:
status: In Progress → Fix Committed
Revision history for this message
Johannes Ring (johannr) wrote :

The submesh demo is disabled because of this bug. Should it be enabled again now that this bug has been fixed? I tried to run it but there were lots of nan's in the smoothed mesh, which is the same as reported in bug #1143977.

Revision history for this message
Jan Blechta (blechta) wrote : Re: [Bug 1047641] Re: HarmonicSmoothing::move is broken

On Tue, 12 Mar 2013 13:54:43 -0000
Johannes Ring <email address hidden> wrote:
> The submesh demo is disabled because of this bug. Should it be enabled
> again now that this bug has been fixed? I tried to run it but there
> were lots of nan's in the smoothed mesh, which is the same as
> reported in bug #1143977.
>

That's different functionality: HarmonicSmoothing vs MeshSmoothing.
Just remove fluid_mesh.smooth() in demo. Still fluid_mesh is smoothed by
HarmonicSmoothing on line fluid_mesh.move(structure_mesh).

Jan

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.