dolfin team mailing list archive

[dolfin@xxxxxxxxxx (DOLFIN)]

One or more new changesets pushed to the primary DOLFIN repository.
A short summary of the last three changesets is included below.

changeset:   2452:8b6dfddc6ca93195d961547b3ad498bc73aeeba0
tag:         tip
parent:      2451:f1d485299d95732734d29f8331a5d858041f8ec2
parent:      2450:3356dec0ee79e29cd152e963e7d70ff56e785081
user:        "Anders Logg <logg@xxxxxxxxx>"
date:        Wed Nov 22 14:09:02 2006 +0100
files:       
description:
Merge


changeset:   2451:f1d485299d95732734d29f8331a5d858041f8ec2
parent:      2448:de52acef02c04ea6705c3b562e0f3925d614fd5d
user:        "Anders Logg <logg@xxxxxxxxx>"
date:        Wed Nov 22 13:16:10 2006 +0100
files:       src/bench/fem/assembly/AdvectionOperator_3D_3_FER.form src/bench/fem/assembly/AdvectionOperator_3D_3_FER.h src/bench/fem/assembly/AdvectionOperator_3D_3_FFC.form src/bench/fem/assembly/AdvectionOperator_3D_3_FFC.h src/bench/fem/assembly/Makefile src/bench/fem/assembly/README src/bench/fem/assembly/bench.log src/bench/fem/assembly/bench.py src/bench/fem/assembly/main.cpp src/kernel/la/PETScMatrix.cpp
description:
Add benchmark for assembly.


changeset:   2450:3356dec0ee79e29cd152e963e7d70ff56e785081
user:        "Johan Jansson <johanjan@xxxxxxxxxxxxxxxx>"
date:        Wed Nov 22 13:12:05 2006 +0100
files:       src/demo/scripting/pydolfin/solvers/elasticity-updated/direct/directpde.py src/demo/scripting/pydolfin/solvers/elasticity-updated/direct/problem.py src/kernel/ode/MonoAdaptiveFixedPointSolver.cpp src/kernel/ode/dolfin/MonoAdaptiveFixedPointSolver.h src/kernel/parameter/dolfin/DefaultParameters.h
description:
ODE:

Implemented optional experimental fixed-point stabilization for the
mono-adaptive case. The stabilization is not yet automated, and is
controlled by parameters:

"ODE fixed-point stabilize"
"ODE fixed-point stabilization m"
"ODE fixed-point stabilization l"
"ODE fixed-point stabilization ramp"

Need to adaptively/automatically compute these.

PyDOLFIN:

Testing stabilization in the elasticity-updated/direct demo. Looks
promising, a factor 8 (in f() evaluations, should be the same in
execution time) improvement over the stable limit for this problem.

Limited comparison against Newton/Krylov also wins with a factor 4-8
for this problem.


----------------------------------------------------------------------
For more details, visit http://www.fenics.org/hg/dolfin