Klaus Burkart <k_burk...@yahoo.com> writes: > For a symmetric matrix: > I assume, I should compute the number of nonzeros per row just for the upper > triangle?
Yes. > > Am Montag, 16. April 2018, 14:23:19 MESZ hat Jed Brown > <j...@jedbrown.org> Folgendes geschrieben: > > Klaus Burkart <k_burk...@yahoo.com> writes: > >> Hi, >> How can I preallocate space for matrices (some symmetric, others >> asymmetric) if I have the global number of nonzeros (NNZ) but not the >> number of nonzeros per row? > > Unfortunately, this is not enough for efficient assembly. It is more > efficient to over-estimate. > >> I could compute the NNZ for the upper or lower part separately if this >> would be useful for symmetric matrices. >> >> I create the matrix this way: >> MatCreate(PETSC_COMM_WORLD, &M); >> MatSetSizes(M, PETSC_DECIDE, PETSC_DECIDE, N, N); >> MatSetFromOptions(M); >> Klaus