Matrix: Dziekonski/dielFilterV2clx
Description: High-order vector finite element method in EM
(undirected graph drawing) |
Matrix properties | |
number of rows | 607,232 |
number of columns | 607,232 |
nonzeros | 25,309,272 |
structural full rank? | yes |
structural rank | 607,232 |
# of blocks from dmperm | 1 |
# strongly connected comp. | 1 |
explicit zero entries | 0 |
nonzero pattern symmetry | symmetric |
numeric value symmetry | 97% |
type | complex |
structure | symmetric |
Cholesky candidate? | no |
positive definite? | no |
author | A. Dziekonski, A. Lamecki, M. Mrozowski |
editor | T. Davis |
date | 2011 |
kind | electromagnetics problem |
2D/3D problem? | yes |
Additional fields | size and type |
b | full 607232-by-1 |
Notes:
High order vector finite element method in electromagnetics The dielFilter* matrices came from analysis of a 4th-pole dielectric resonator [4] generated with Finite Element Method. The tetrahedral mesh of the structure was generated with the Netgen mesher [2]. The matrices were used as an example in our paper [3]. dielFilterV2clx - complex symmetric matrix (607,232 x 607,232), 25,309,272 nonzero (real) and 728,900 nonzero (imag) elements. First 109,108 unknowns correspond to lowest level base functions. dielFilterV2real - real symmetric matrix (1,157,456 x 1,157,456) and 48,538,952 nonzero elements. First 209,432 unknowns correspond to lowest level base functions. dielFilterV3clx - complex symmetric matrix (420,408 x 420,408), 32,886,208 nonzero (real) and 3,706,513 (imag) elements. First 24,716 unknowns correspond to lowest level base functions, next 116,152 unknowns correspond to the second level. dielFilterV3real - real symmetric matrix (1,102,824 x 1,102,824) and 89,306,020 nonzero elements. First 66,353 unknowns correspond to lowest level base functions, next 305,729 unknowns correspond to the second level. All matrices are sparse and come with right-hand-sides. [2] J. Schoberl, "NETGEN An advancing front 2D/3D-mesh generator based on abstract rules," Computing and Visualization in Science, vol. 1, No. 1, pp. 41-52, July 1997 [3] A. Dziekonski, A. Lamecki, M. Mrozowski, Tuning A Hybrid GPU-CPU V-cycle Multilevel Preconditioner for Solving Large Real and Complex Systems of FEM Equations. [4] F. Alessandri, M. Chiodetti, A. Giugliarelli; D. Maiarelli, G. Martirano, D. Schmitt, L. Vanni and F. Vitulli. The electric-field Integral-equation method for the analysis and design of a class of rectangular cavity filters loaded by dielectric and metallic cylindrical pucks, Microwave Theory and Techniques, IEEE Transactions on, vol. 52, no 8, pp. 1790-1797, Aug. 2004.
Ordering statistics: | result |
nnz(chol(P*(A+A'+s*I)*P')) with AMD | 426,850,000 |
Cholesky flop count | 1.3e+12 |
nnz(L+U), no partial pivoting, with AMD | 853,092,768 |
nnz(V) for QR, upper bound nnz(L) for LU, with COLAMD | 1,187,745,302 |
nnz(R) for QR, upper bound nnz(U) for LU, with COLAMD | 2,848,515,948 |
For a description of the statistics displayed above, click here.
Maintained by Tim Davis, last updated 12-Mar-2014.
Matrix pictures by cspy, a MATLAB function in the CSparse package.
Matrix graphs by Yifan Hu, AT&T Labs Visualization Group.