Hi Timo, Am Donnerstag, 7. September 2017 13:40:19 UTC+2 schrieb Timo Heister: > > Korinna, > > > 1) Maybe I found a wrong assertion in the FETools. > > I have trouble with FETools while working in the debug mode. I > cannot > > use the provided FETools::interpolation for the higher order > reconstruction, > > It might be that we haven't been testing this function enough, so a > bug like this is not unlikely. > > > The violated condition was: > > owned_elements.size()==size() > > Additional information: > > The locally owned elements have not been properly initialized! > > This happens for example if this object has been initialized with > exactly > > one overlapping IndexSet. > > Can you try initializing your ghosted vector with > vec.reinit(locally_owned_index_set, locally_relevant_index_set, mpi_com); > instead of only specifying the ghosted set? Maybe that fixes the > exception. >
Thanks for great tipp. That was rather easy. But should I use the locally_relevant_index_set as the second argument or locally_relevant_dofs.subtract_set(locally_owned_dofs)? Both versions work for me so far. > > 2) I have also a much worse problem right now, for which I have not > found a > > satisfying solution yet. > > It seems to me that I cannot use the MeshSmooting methods for the > > parallel::distributed::Triangulation. In particular the flag > patch_level_1 > > causes trouble. > > Yes, mesh smoothing in parallel is a problem at the moment. It is > basically not working at all. Sorry. I was hoping to have an > implementation for that soon (we did some progress towards it last > month). See > https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_dealii_dealii_pull_1711&d=DwIBaQ&c=Ngd-ta5yRYsqeUsEDgxhcqsYYY1Xs5ogLxWPA_2Wlc4&r=4k7iKXbjGC8LfYxVJJXiaYVu6FRWmEjX38S7JmlS9Vw&m=fx5zPVPj5YHLwSgXHBKNn1I9OvTImEnlVgDaDmSH-Z0&s=xMjuyhQn_tZbaM_K32G208xiQJ7EjMmKAmSRfDoy_4Y&e= > > > https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_dealii_dealii_issues_1655&d=DwIBaQ&c=Ngd-ta5yRYsqeUsEDgxhcqsYYY1Xs5ogLxWPA_2Wlc4&r=4k7iKXbjGC8LfYxVJJXiaYVu6FRWmEjX38S7JmlS9Vw&m=fx5zPVPj5YHLwSgXHBKNn1I9OvTImEnlVgDaDmSH-Z0&s=HmN6gJQSWE1WxMwrHI5csg3_TaC8JnCXyOa6GAnnxGQ&e= > > and issues linked in > there. > One workaround for now might be to switch to a shared::Triangulation. > Oh. Okay. Thank you for the information. Sorry for the duplicate thread. I will read those threads carefully next. So I guess it is not possible right now to only prepare and excecute refinement without considering any coarsening flags? Best, Korinna -- 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 [email protected]. For more options, visit https://groups.google.com/d/optout.
