On 04/19/2018 07:15 AM, 'Maxi Miller' via deal.II User Group wrote:
| std::vector<bool>TE_value_mask(dof_handler.n_locally_owned_dofs()),
TL_value_mask(dof_handler.n_locally_owned_dofs()),
N_value_mask(dof_handler.n_locally_owned_dofs());
DoFTools::extract_dofs(dof_handler,surface_N_mask,N_value_mask);
DoFTools::extract_dofs(dof_handler,surface_TE_mask,TE_value_mask);
DoFTools::extract_dofs(dof_handler,surface_TL_mask,TL_value_mask);
LinearAlgebraTrilinos::MPI::Vectorunified_test_vector(dof_handler.locally_owned_dofs(),mpi_communicator);
for(autoindex :scaled_solution.locally_owned_elements())
{
if(N_value_mask[index])
{
unified_test_vector(index)=scaled_solution(index)*(unified_max_N_value
-unified_min_N_value)*unified_N_value +unified_N_value;
}
|
This cannot work (and is likely the cause for your issues):
`N_values_mask` is an array if size `dof_handler.n_locally_owned_dofs()`
whereas `unified_test_vector` is a vector if size
`dof_handler.n_dofs()`. For sequential computations, the two are the
same, but for parallel computations, the two are very different.
What that means is that you cannot expect that `index` is a good index
for both. The way you define `index` is that it iterates over the index
set of locally owned elements (i.e., over indices between zero and
`dof_handler.n_dofs()`), and consequently the statement
N_value_mask[index]
is meaningless. You need to translate between the two.
Best
W.
--
------------------------------------------------------------------------
Wolfgang Bangerth email: [email protected]
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 [email protected].
For more options, visit https://groups.google.com/d/optout.
