Hello,
I am working on modeling a phenomenon that is similar to Stokes flow in and
around a droplet. In my model there are two sets of Stokes equations that
govern the flow in each one of the media that are made of different
materials – one set for the flow in the droplet, and the other f
Your suggestions worked exactly as advertised. I've copy-pasta's my
updated code below for completeness.
> Thomas,
> your code looks essentially correct, though it can be improved:
>
> > Now, for my attempt at this:
> >
> > |
> > template(int spacedim>
> > void data_dot_normal (Vector
> I must reinstall Petsc and deal.ii again. Is there a need to reinstall P4est
> as well ?
p4est also uses MPI so yes.
By the way: after you install deal.II, you should run "make test" in
the build directory. I am pretty sure it would have detected this
problem.
--
Timo Heister
http://www.math.
Respected Professor,
Yes, I remember this discussion regarding not using --download mpich tag
while installing Petsc happening on github as well.
Yes, I think I should now do it. Although it still remains a mystery to me,
how and why openmpi got installed and is now clashing with mpich.
Anyway
On 10/28/2016 11:49 AM, RAJAT ARORA wrote:
I think that openmpi and Mpich are clashing now. (I didn't even know
that both were installed upto now).
I think, if I can disable openmpi somehow, I will be in good shape. This
is because, I compiled deal.ii with mpich library which was installed by
p
I am surprised why it is not working. I can't recall what has changed.
I don't remember installing any new libraries. It was working till Monday.
Also, I have installed petsc with --download mpich tag but donot have any
other mpi installation.
and more importantly, it was working until monday.
I also played with bash for windows last week and got the same
conclusion: it is super easy to get it to work. I was too busy to
write up an FAQ entry, but I agree that this is probably the best
option on windows right now.
> I would guess that one can also just use Eclipse, which is free and wil
>> Should I reinstall my mpi library ? Does deal.ii will also need to be
>> recompiled then ?
>
> I don't know. You say that it suddenly stopped working. Did it work
> correctly before? What has changed in the meantime?
This can happen if you have conflicting mpi libraries installed (for
example l
On 10/28/2016 09:55 AM, RAJAT ARORA wrote:
I am just calling dealii::Utilities::MPI::MPI_InitFinalize
mpi_initialization(argc, argv, 1) once.
I am not making an explicit call to MPI_Init anywhere.
Yes, calling MPI_InitFinalize is enough.
Should I reinstall my mpi library ? Does deal.ii wil
Respected Prof. Bangerth,
I am just calling dealii::Utilities::MPI::MPI_InitFinalize
mpi_initialization(argc, argv, 1) once.
I am not making an explicit call to MPI_Init anywhere.
Should I reinstall my mpi library ? Does deal.ii will also need to be
recompiled then ?
On Friday, October 28, 2
On 10/28/2016 09:47 AM, RAJAT ARORA wrote:
Yes, you are right :).
When running with 4 mpi processes, all threads print their rank as 0.
Why is it broken all of a sudden. What is this problem called ? How do I
correct it ?
I don't know. Are you calling MPI_Init, for example?
W.
--
---
Respected Prof. Bangerth,
Yes, you are right :).
When running with 4 mpi processes, all threads print their rank as 0.
Why is it broken all of a sudden. What is this problem called ? How do I
correct it ?
On Friday, October 28, 2016 at 11:39:33 AM UTC-4, Wolfgang Bangerth wrote:
>
> On 10/28/20
On 10/24/2016 01:41 PM, benhour.amiria...@gmail.com wrote:
for (typename Triangulation ::active_cell_iterator
cell=triangulation.begin_active();
cell!=triangulation.end(); ++cell)
*for (unsigned int f=0; f::faces_per_cell;
++f)*
*if (cell->face(f)->at_bo
On 10/28/2016 09:34 AM, RAJAT ARORA wrote:
To run the code on n processes, I used to run the command
mpirun -np ./
But, now, when I run it using this command, it runs n programs with n
mpi processes each.
It is like I have executed mpirun -np ./ command n times
and the compiler is then run
Hello all,
I am working on a 3D solid mechanics problem using deal.ii.
To run the code on n processes, I used to run the command
mpirun -np ./
But, now, when I run it using this command, it runs n programs with n mpi
processes each.
It is like I have executed mpirun -np ./ command n times
Hamed,
2016-10-28 10:36 GMT-04:00 Hamed Babaei :
> If I solve a problem with one million DOFs on a cluster of 64 processors,
> every processor will be dealing with around 15000 DOFs.
> So you mean even in that case direct solver would be slow? I thought direct
> solver would solve the problem only
Many thanks Wolfgang and Daneil.
I think Wolfgang's answer is illuminating.
On Friday, October 28, 2016 at 3:13:56 PM UTC+2, Daniel Arndt wrote:
>
> so the uniform flux from left and right of the rectangle implies periodic
>> boundary condition. But the K ( hydraulic conductivity) is a funct
Dear Bruno,
If I solve a problem with one million DOFs on a cluster of 64 processors,
every processor will be dealing with around 15000 DOFs.
So you mean even in that case direct solver would be slow? I thought direct
solver would solve the problem only on those DOFs that it owns.
Best,
Hamed
>
> so the uniform flux from left and right of the rectangle implies periodic
> boundary condition. But the K ( hydraulic conductivity) is a function of
> (x,y).
> If we want to enforce the periodic boundary condition, should we expect to
> have a condition on K ? should K be periodic as well f
On 10/28/2016 06:18 AM, Retired Replicant. wrote:
I am sorry I can not formulate my problem more clearly. This is another try.
so the uniform flux from left and right of the rectangle implies periodic
boundary condition. But the K ( hydraulic conductivity) is a function of (x,y).
If we want to
Hamed,
2016-10-27 19:03 GMT-04:00 Hamed Babaei :
> I've been using the default one, Amesos_KLU, and I checked
> Amesos_Superludist but trilinos doesn't recognize it.
> when we installed dealii and its dependencies on the cluster, I think we
> didn't uncomment the "#PACKAGES="${PACKAGES} once:super
I am sorry I can not formulate my problem more clearly. This is another try.
so the uniform flux from left and right of the rectangle implies periodic
boundary condition. But the K ( hydraulic conductivity) is a function of
(x,y).
If we want to enforce the periodic boundary condition, should we
22 matches
Mail list logo
