Re: boundary region

[Konstantinos Poulios]

Dear Edouard

The function

asm_boundary_qu_term(boundary_num, mim, mf_u, mf_d, q)

accepts a region id as the first argument "boundary_num".

The _expression_ that you have used is for the old (low-level) generic assembly system, which we do not recommend so much anymore.

The function "asm_generic" is now preferred to "asm_boundary". But as said you should be able to use "asm_boundary_qu_term" directly.

Best regards

Kostas

On Sun, Jun 13, 2021 at 7:16 AM Edouard Oudet <edouard.oudet@univ-grenoble-alpes.fr> wrote:

Dear all,

I think I solved my problem by a call like

getfem.asm_boundary(bregion_id, expr, mim, mf, mf, a)

where

  expr = 'a = data(#2); M(#1,#2)+= comp(vBase(#1).vBase(#2).Base(#1))(:,
i,: ,i , j).a(j)'

I am not completely sure to understand the indexing  (:, i,: ,i ,
j).a(j) in previous _expression_..
Best,
Edouard.




On 6/12/21 8:39 PM, Edouard Oudet wrote:
> Dear Getfem users,
> It seems that the python function `asm_boundary_qu_term` does not
> accept a (boundary faces) region argument.
> Is there a way to compute the qu matrix only considering a sublist of
> boundary faces (/ segments) from python?
> Thanks,
> best,
> Edouard.
>

--
Edouard Oudet : http://www-ljk.imag.fr/membres/Edouard.Oudet/
IMAG - Bureau 164
700 avenue Centrale
38400 Saint Martin d'Hères
+33 (0)4 57 42 17 71  (office LJK)