Thank you for your reply. The recursive idea looks interesting. I will try
that for simple demonstratable cases. Also as a side question, p4est is
still relevant today right ? Are there any disadvantages of using it with
deal.ii at either very large or moderately large scales ?
Thanks,
Pratik.
On Friday, April 27, 2018 at 12:41:41 AM UTC-4, Wolfgang Bangerth wrote:
>
> On 04/26/2018 03:04 PM, Pratik wrote:
> >
> > In this function, there are two for loops (line 64 to 74 and line 84 to
> 97).
> > First loop returns the ghost cell and gets the subdomain_id of that cell
> into
> > the sc_array subids and the second loop adds this subdomain_id
> information to
> > the vertices_with_ghost_neighbors.
> >
> > I was wondering if it possible to also do something like this in the
> first
> > loop as well:
> >
> > |
> > for(j =0;j <nsubs;j++)
> > {
> > (*vertices_with_ghost_neighbors)[cell->vertex_index(sides[i].corner)]
> > .insert (subdomain_ids[j]);
> > }
> > |
> >
> >
> > so that even the ghost cells to the first layer of ghost cells are added
> to
> > the vertices_with_ghost_neighbors. Would this work ? Is there a better
> way ?
>
> I have to admit that I don't know what the function does, exactly, and
> that
> it's a bit hard to read -- we definitely had a different programming style
> at
> the time :-(
>
> What happens if you try this? The way to test these sorts of things is to
> come
> up with a simple case for which you can draw little pictures -- say, you
> start
> with a unit square and then refine it 2 or 3 times. Run things with two
> processors and print a lot of debug output (or run things in a debugger)
> --
> you can follow each step this way, and with a piece of paper and a pencil,
> you
> can draw little pictures that help you understand what is supposed to
> happen
> and what actually happens.
>
>
> I apologize for not being of more help. The code was written some ten
> years
> ago by Timo Heister when he was visiting my lab. I have no recollection of
> how
> this code works, and would have to spend significant time to read through
> it
> to be better able to help -- time I unfortunately don't have. Timo might
> remember more, but he's got other things going on in his life right now as
> well.
>
> Best
> W.
>
> --
> ------------------------------------------------------------------------
> Wolfgang Bangerth email: bang...@colostate.edu
> <javascript:>
> 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.
