1KSP Object: (physical_) 4 MPI processes 2 type: cg 3 maximum iterations=10, initial guess is zero 4 tolerances: relative=1e-05, absolute=1e-50, divergence=10000. 5 left preconditioning 6 using PRECONDITIONED norm type for convergence test 7PC Object: (physical_) 4 MPI processes 8 type: bddc 9 Use verbose output: 0 10 Use user-defined CSR: 1 11 Use local mat graph: 0 12 Connectivity graph topological dimension: 2 13 Use vertices: 1 (vertex size 1) 14 Use edges: 1 15 Use faces: 0 16 Use true near null space: 0 17 Use QR for single constraints on cc: 0 18 Use change of basis on local edge nodes: 0 19 Use change of basis on local face nodes: 0 20 User defined change of basis matrix: 0 21 Has change of basis matrix: 0 22 Eliminate dirichlet boundary dofs: 0 23 Switch on static condensation ops around the interface preconditioner: 0 24 Use exact dirichlet trick: 1 25 Interface extension: DIRICHLET 26 Multilevel max levels: 1 27 Multilevel coarsening ratio: 4 28 Use estimated eigs for coarse problem: 0 29 Use deluxe scaling: 0 30 Use deluxe zerorows: 0 31 Use deluxe singlemat: 0 32 Rebuild interface graph for Schur principal minors: 0 33 Number of dofs' layers for the computation of principal minors: -1 34 Use user CSR graph to compute successive layers: 0 35 Adaptive constraint selection threshold (active 0, userdefined 0): 0. 36 Min constraints / connected component: 0 37 Max constraints / connected component: 0 38 Invert exact Schur complement for adaptive selection: 0 39 Symmetric computation of primal basis functions: 1 40 Num. Procs. to map coarse adjacency list: 0 41 Coarse eqs per proc (significant at the coarsest level): 1 42 Detect disconnected: 0 (filter 0) 43 Benign subspace trick: 0 (change explicit 0) 44 Benign subspace trick is active: 0 45 Algebraic computation of no-net-flux: 0 46 ********************************** STATISTICS AT LEVEL 0 ********************************** 47 Global dofs sizes: all 342 interface 36 coarse 9 48 Coarsening ratios: all/coarse 38 interface/coarse 4 49 Active processes : 4 50 Total subdomains : 4 51 Dofs type : MIN MAX MEAN 52 Interior dofs : 72 81 76 53 Interface dofs : 18 19 18 54 Primal dofs : 5 5 5 55 Local dofs : 90 100 95 56 Local subs : 1 1 57 --- Interior solver (rank 0) 58 KSP Object: (physical_pc_bddc_dirichlet_) 1 MPI processes 59 type: preonly 60 maximum iterations=10000, initial guess is zero 61 tolerances: relative=1e-05, absolute=1e-50, divergence=10000. 62 left preconditioning 63 using NONE norm type for convergence test 64 PC Object: (physical_pc_bddc_dirichlet_) 1 MPI processes 65 type: cholesky 66 out-of-place factorization 67 tolerance for zero pivot 2.22045e-14 68 matrix ordering: natural 69 factor fill ratio given 5., needed 3.48 70 Factored matrix follows: 71 Mat Object: 1 MPI processes 72 type: seqsbaij 73 rows=72, cols=72 74 package used to perform factorization: petsc 75 total: nonzeros=1044, allocated nonzeros=1044 76 block size is 1 77 linear system matrix = precond matrix: 78 Mat Object: (physical_pc_bddc_dirichlet_) 1 MPI processes 79 type: seqsbaij 80 rows=72, cols=72 81 total: nonzeros=300, allocated nonzeros=300 82 total number of mallocs used during MatSetValues calls=0 83 block size is 1 84 --- Correction solver (rank 0) 85 KSP Object: (physical_pc_bddc_neumann_) 1 MPI processes 86 type: preonly 87 maximum iterations=1000, initial guess is zero 88 tolerances: relative=1e-10, absolute=1e-10, divergence=10000. 89 left preconditioning 90 using NONE norm type for convergence test 91 PC Object: (physical_pc_bddc_neumann_) 1 MPI processes 92 type: lu 93 out-of-place factorization 94 tolerance for zero pivot 2.22045e-14 95 matrix ordering: nd 96 factor fill ratio given 5., needed 2.86706 97 Factored matrix follows: 98 Mat Object: 1 MPI processes 99 type: seqaij 100 rows=87, cols=87 101 package used to perform factorization: petsc 102 total: nonzeros=1941, allocated nonzeros=1941 103 not using I-node routines 104 linear system matrix = precond matrix: 105 Mat Object: (physical_pc_bddc_neumann_) 1 MPI processes 106 type: seqaij 107 rows=87, cols=87 108 total: nonzeros=677, allocated nonzeros=677 109 total number of mallocs used during MatSetValues calls=0 110 not using I-node routines 111 --- Coarse solver 112 KSP Object: (physical_pc_bddc_coarse_) 1 MPI processes 113 type: preonly 114 maximum iterations=1000, initial guess is zero 115 tolerances: relative=1e-10, absolute=1e-10, divergence=10000. 116 left preconditioning 117 using NONE norm type for convergence test 118 PC Object: (physical_pc_bddc_coarse_) 1 MPI processes 119 type: redundant 120 First (color=0) of 1 PCs follows 121 KSP Object: (physical_pc_bddc_coarse_redundant_) 1 MPI processes 122 type: preonly 123 maximum iterations=10000, initial guess is zero 124 tolerances: relative=1e-05, absolute=1e-50, divergence=10000. 125 left preconditioning 126 using NONE norm type for convergence test 127 PC Object: (physical_pc_bddc_coarse_redundant_) 1 MPI processes 128 type: lu 129 out-of-place factorization 130 tolerance for zero pivot 2.22045e-14 131 matrix ordering: nd 132 factor fill ratio given 5., needed 1.12308 133 Factored matrix follows: 134 Mat Object: 1 MPI processes 135 type: seqaij 136 rows=9, cols=9 137 package used to perform factorization: petsc 138 total: nonzeros=73, allocated nonzeros=73 139 using I-node routines: found 3 nodes, limit used is 5 140 linear system matrix = precond matrix: 141 Mat Object: (physical_pc_bddc_coarse_) 1 MPI processes 142 type: seqaij 143 rows=9, cols=9 144 total: nonzeros=65, allocated nonzeros=65 145 total number of mallocs used during MatSetValues calls=0 146 using I-node routines: found 5 nodes, limit used is 5 147 linear system matrix = precond matrix: 148 Mat Object: (physical_pc_bddc_coarse_) 1 MPI processes 149 type: seqaij 150 rows=9, cols=9 151 total: nonzeros=65, allocated nonzeros=65 152 total number of mallocs used during MatSetValues calls=0 153 using I-node routines: found 5 nodes, limit used is 5 154 linear system matrix = precond matrix: 155 Mat Object: 4 MPI processes 156 type: is 157 rows=342, cols=342 158 total: nonzeros=1724, allocated nonzeros=36200 159 total number of mallocs used during MatSetValues calls=0 160 has attached near null space 161---------------------BDDC stats------------------------------- 162Number of degrees of freedom : 342 163Eigenvalues preconditioned operator : 1.0e+00 1.3e+00 164-------------------------------------------------------------- 165------------------FETI-DP stats------------------------------- 166Number of degrees of freedom : 31 167Eigenvalues preconditioned operator : 1.0e+00 1.3e+00 168-------------------------------------------------------------- 169