Re: [deal.II] Usage of MeshWorker::mesh_loop for DG methods

[Wolfgang Bangerth]

On 1/25/19 3:44 AM, 'Maxi Miller' via deal.II User Group wrote:
> When using mesh_loop for discontinuous galerkin, I have to add two additional 
> functions, boundary_worker and face_worker. When calling face_worker, it gets 
> two cell-iterators as parameter. Are those two cells automatically 
> neighbours, 
> or do I have to check that in the worker itself? I could not find that in the 
> documentation.

Yes, these are the two adjacent cells, since you need this information when 
computing jump terms.

Would you be interested in adding this information to whatever place you 
thought should have said this? (I.e., where you looked for it?)

Best
  W.

-- 

Wolfgang Bangerth  email: bange...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [deal.II] output vector solution to vtk file

[Phạm Ngọc Kiên]

I really appreciate your answer. The library is really useful and i will
dig deeper into it.
I am using the latest version of dealii.
Thank you very much.
Best regards.
Pham Ngoc Kien

Vào 04:42 PM, T.6, 25 Th1, 2019 Jean-Paul Pelteret 
đã viết:

> Dear Pham,
>
> Welcome to deal.II!
>
> Your observation is related to this frequently asked question
> .
> Basically, deal.II outputs the solution on each cell individually because
> we cater for the situation where the solution is discontinuous between
> elements. This could arise when the solution field is discretised by
> discontinuous finite elements, when there are hanging nodes in the
> triangulation, or when a post-processed quantity is naturally discontinuous
> between elements. This patch-based output, although maybe a bit wasteful in
> the case of continuous fields due to the data duplication, is the most
> robust solution for all scenarios.
>
> For that particular file type and the default output options, the data
> points are located at the vertices of the triangulation. If you are using
> the developer version of deal.II it is now possible to output the solution
> to "higher-order cells” (e.g. quadratic or cubic interpolatory cells). In
> this case, you’d get extra data points on the edges and interior of each
> cell.
>
> I hope that this helped!
>
> Best,
> Jean-Paul
>
> On 25 Jan 2019, at 07:35, Phạm Ngọc Kiên 
> wrote:
>
> Dear Prof. Wolfgang Bangerth,
> I am a newbie with deal.II library, and I have a question in this topic.
> Thus, I put my question here.
> I can now output my vector values results using DataPostprocessor class
> into vtk file.
> However, when I opened the output file, I see some data points in the file
> had the same positions. I had checked with the examples (step-3) and
> received the same thing.
> Could you please tell me why it is?
> In my limited opinion, the function: attach_dof_handler(dof_handler) will
> give me the data positions which are the positions of all nodes in my mesh.
> I wonder if the same data points I mentioned above are the nodes on edge,
> cell, face?
> I would like to thank you very much in advance.
>
> Best regards,
> Pham Ngoc Kien
>
> --
> The deal.II project is located at http://www.dealii.org/
> For mailing list/forum options, see
> https://groups.google.com/d/forum/dealii?hl=en
> ---
> You received this message because you are subscribed to the Google Groups
> "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to dealii+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>
>
> --
> The deal.II project is located at http://www.dealii.org/
> For mailing list/forum options, see
> https://groups.google.com/d/forum/dealii?hl=en
> ---
> You received this message because you are subscribed to the Google Groups
> "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to dealii+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[deal.II] Re: Trouble Installing: "undefined symbol: cblas_ctrmv"

[kylew]

I see now that the symbols are actually there, but they are dynamic symbols 
which don't seem to get picked up (Though I don't really know what the 
difference is).

-Kyle

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[deal.II] Usage of MeshWorker::mesh_loop for DG methods

['Maxi Miller' via deal.II User Group]

When using mesh_loop for discontinuous galerkin, I have to add two 
additional functions, boundary_worker and face_worker. When calling 
face_worker, it gets two cell-iterators as parameter. Are those two cells 
automatically neighbours, or do I have to check that in the worker itself? 
I could not find that in the documentation.
Thanks!

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[deal.II] Re: Trouble Installing: "undefined symbol: cblas_ctrmv"

[Denis Davydov]

Hi Kyle

On Friday, January 25, 2019 at 9:11:35 AM UTC+1, ky...@math.uh.edu wrote:
>
> Hi Denis,
>
> I did see both of the issues raised on github before I posted, but it 
> wasn't really clear what the resolution was other than its a gsl dependency 
> problem. 
>

I would say from our side we tried to pick up `gslcblas` to fix part of the 
issue with Linux provided GSL
 

>
> nm -g /usr/lib64/libgslcblas.so
> nm -g /usr/lib64/libgsl.so
> both tell me there are no symbols, even without grepping for the one in 
> question.
>
> Does this mean that if I do need GSL I should install from source instead? 
> (as pointed out in the other response, I don't need GSL for the work i will 
> be doing).
>

Yes, I would say so. But given that you won't need this functionality, you 
can safely disable GSL as recommended by David. 
In recent years I tend to prefer build everything from sources 
using https://github.com/dealii/dealii/wiki/deal.II-in-Spack for all 
environments (macOS and HPC).

Regards,
Denis.
 

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[deal.II] Re: Trouble Installing: "undefined symbol: cblas_ctrmv"

[kylew]

Hi Denis,

I did see both of the issues raised on github before I posted, but it 
wasn't really clear what the resolution was other than its a gsl dependency 
problem. 

nm -g /usr/lib64/libgslcblas.so
nm -g /usr/lib64/libgsl.so
both tell me there are no symbols, even without grepping for the one in 
question.

Does this mean that if I do need GSL I should install from source instead? 
(as pointed out in the other response, I don't need GSL for the work i will 
be doing).

-Kyle

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.