[petsc-users] Question about DMDAGetElements

[Sajid Ali]

Hi,

I'm trying to understand this example
<https://github.com/haplav/petsc-tut-it4i-2018/blob/master/exercises/ex6.c>
from a tutorial. DM is used as an alternative to setting up the matrix and
vector separately (as was done in the previous example
<https://github.com/haplav/petsc-tut-it4i-2018/blob/master/exercises/ex5.c>).


Since DMDACreate1d was used to create the mesh, can someone explain this
logic more clearly :

  /*
    Gets an array containing the indices (in local coordinates)
    of all the local elements.
    nel	- number of local elements
    nen	- number of one element's nodes
    e	- the local indices of the elements' vertices
  */
  ierr = DMDAGetElements(da, &nel, &nen, &e);CHKERRQ(ierr);

  /*
     Assemble matrix and RHS
  */
  value[0] = 1.0; value[1] = -1.0; value[2] = -1.0; value[3] = 1.0;
  bvalue[0] = 1.0; bvalue[1] = 1.0;
  for (i=0; i<nel; i++) {
    ierr = MatSetValuesLocal(A, 2, e+nen*i, 2, e+nen*i, value,
ADD_VALUES);CHKERRQ(ierr);
    //TODO use VecSetValuesLocal() to set the values
  }
  ierr = DMDARestoreElements(da, &nel, &nen, &e);CHKERRQ(ierr);


The dm was created using

  ierr = MPI_Allreduce(&n, &N, 1, MPI_INT, MPI_SUM,
PETSC_COMM_WORLD);CHKERRQ(ierr);
  ierr = DMDACreate1d(PETSC_COMM_WORLD,DM_BOUNDARY_NONE,N,1,1,NULL,&da);CHKERRQ(ierr);


Using n = 3 and I get the following results when I print the values of
nel,nen and e using this statement
PetscPrintf(PETSC_COMM_SELF,"nel : %d,nen : %d, e elements :
%d,%d,%d,%d,%d,%d,%d,%d ,rank
%d\n",nel,nen,e[0],e[1],e[2],e[3],e[4],e[5],e[6],e[7],rank);

The output comes out as  :

[sajid at xrm exercises]$ mpirun -np 4 ./ex6
nel : 2,nen : 2, e elements : 0,1,1,2,1936617321,0,3,0 ,rank 0
nel : 3,nen : 2, e elements : 0,1,1,2,2,3,4,0 ,rank 1
nel : 3,nen : 2, e elements : 0,1,1,2,2,3,4,0 ,rank 2
nel : 3,nen : 2, e elements : 0,1,1,2,2,3,3,0 ,rank 3

Does this mean that I'm supposed to read the first 4 elements for rank 0 as
row0, col0, row1, col1?

And how does differ If I'm creating a vector using DMCreate1d ?


Thank You,
Sajid Ali
Applied Physics
Northwestern University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20181112/7e35edbb/attachment.html>