Use LOBPCG with Epetra, for a generalized eigenvalue problem.
Use LOBPCG with Epetra, for a generalized eigenvalue problem.This example computes the eigenvalues of largest magnitude of an generalized eigenvalue problem, using Anasazi's implementation of the LOBPCG method, with Epetra linear algebra.
#include "Epetra_CrsMatrix.h"
#include "Teuchos_CommandLineProcessor.hpp"
#include "Teuchos_StandardCatchMacros.hpp"
#ifdef HAVE_MPI
#include "Epetra_MpiComm.h"
#include <mpi.h>
#else
#include "Epetra_SerialComm.h"
#endif
#include "Epetra_Map.h"
#include "ModeLaplace2DQ2.h"
int main(int argc, char *argv[]) {
#ifdef HAVE_MPI
MPI_Init(&argc,&argv);
#endif
bool success = false;
try {
#ifdef HAVE_MPI
Epetra_MpiComm Comm(MPI_COMM_WORLD);
#else
Epetra_SerialComm Comm;
#endif
BasicOutputManager<double> printer;
printer.stream(Errors) << Anasazi_Version() << std::endl << std::endl;
std::string which("SM");
Teuchos::CommandLineProcessor cmdp(false,true);
cmdp.setOption("sort",&which,"Targetted eigenvalues (SM or LM).");
if (cmdp.parse(argc,argv) != Teuchos::CommandLineProcessor::PARSE_SUCCESSFUL) {
throw -1;
}
typedef Epetra_MultiVector MV;
typedef Epetra_Operator OP;
typedef MultiVecTraits<double, Epetra_MultiVector> MVT;
const int space_dim = 2;
std::vector<double> brick_dim( space_dim );
brick_dim[0] = 1.0;
brick_dim[1] = 1.0;
std::vector<int> elements( space_dim );
elements[0] = 10;
elements[1] = 10;
Teuchos::RCP<ModalProblem> testCase =
Teuchos::rcp( new ModeLaplace2DQ2(Comm, brick_dim[0], elements[0], brick_dim[1], elements[1]) );
Teuchos::RCP<Epetra_CrsMatrix> K = Teuchos::rcp( const_cast<Epetra_CrsMatrix *>(testCase->getStiffness()), false );
Teuchos::RCP<Epetra_CrsMatrix> M = Teuchos::rcp( const_cast<Epetra_CrsMatrix *>(testCase->getMass()), false );
int nev = 10;
int blockSize = 5;
int maxIters = 500;
double tol = 1.0e-8;
Teuchos::RCP<Epetra_MultiVector> ivec = Teuchos::rcp( new Epetra_MultiVector(K->OperatorDomainMap(), blockSize) );
ivec->Random();
Teuchos::RCP<BasicEigenproblem<double, MV, OP> > MyProblem =
Teuchos::rcp( new BasicEigenproblem<double, MV, OP>(K, M, ivec) );
MyProblem->setHermitian(true);
MyProblem->setNEV( nev );
bool boolret = MyProblem->setProblem();
if (boolret != true) {
printer.print(Errors,"Anasazi::BasicEigenproblem::setProblem() returned an error.\n");
throw -1;
}
Teuchos::ParameterList MyPL;
MyPL.set( "Which", which );
MyPL.set( "Block Size", blockSize );
MyPL.set( "Maximum Iterations", maxIters );
MyPL.set( "Convergence Tolerance", tol );
MyPL.set( "Full Ortho", true );
MyPL.set( "Use Locking", true );
MyPL.set( "Verbosity", verbosity );
LOBPCGSolMgr<double, MV, OP> MySolverMan(MyProblem, MyPL);
Eigensolution<double,MV> sol = MyProblem->getSolution();
std::vector<Value<double> > evals = sol.Evals;
Teuchos::RCP<MV> evecs = sol.Evecs;
std::vector<double> normR(sol.numVecs);
if (sol.numVecs > 0) {
Teuchos::SerialDenseMatrix<int,double> T(sol.numVecs, sol.numVecs);
Epetra_MultiVector Kvec( K->OperatorDomainMap(), evecs->NumVectors() );
Epetra_MultiVector Mvec( M->OperatorDomainMap(), evecs->NumVectors() );
T.putScalar(0.0);
for (int i=0; i<sol.numVecs; i++) {
T(i,i) = evals[i].realpart;
}
K->Apply( *evecs, Kvec );
M->Apply( *evecs, Mvec );
MVT::MvTimesMatAddMv( -1.0, Mvec, T, 1.0, Kvec );
MVT::MvNorm( Kvec, normR );
}
std::ostringstream os;
os.setf(std::ios_base::right, std::ios_base::adjustfield);
os<<"Solver manager returned " << (returnCode == Converged ? "converged." : "unconverged.") << std::endl;
os<<std::endl;
os<<"------------------------------------------------------"<<std::endl;
os<<std::setw(16)<<"Eigenvalue"
<<std::setw(18)<<"Direct Residual"
<<std::endl;
os<<"------------------------------------------------------"<<std::endl;
for (int i=0; i<sol.numVecs; i++) {
os<<std::setw(16)<<evals[i].realpart
<<std::setw(18)<<normR[i]/evals[i].realpart
<<std::endl;
}
os<<"------------------------------------------------------"<<std::endl;
printer.print(Errors,os.str());
success = true;
}
TEUCHOS_STANDARD_CATCH_STATEMENTS(true, std::cerr, success);
#ifdef HAVE_MPI
MPI_Finalize();
#endif
return ( success ? EXIT_SUCCESS : EXIT_FAILURE );
}
Basic implementation of the Anasazi::Eigenproblem class.
Basic output manager for sending information of select verbosity levels to the appropriate output str...
Anasazi header file which uses auto-configuration information to include necessary C++ headers.
Declarations of Anasazi multi-vector and operator classes using Epetra_MultiVector and Epetra_Operato...
The Anasazi::LOBPCGSolMgr provides a powerful solver manager for the LOBPCG eigensolver.
Namespace Anasazi contains the classes, structs, enums and utilities used by the Anasazi package.
ReturnType
Enumerated type used to pass back information from a solver manager.