mfem::PowerMethod Class Reference

PowerMethod helper class to estimate the largest eigenvalue of an operator using the iterative power method. More...

#include <operator.hpp>

Public Member Functions
	PowerMethod ()
	PowerMethod ()
	PowerMethod (MPI_Comm comm_)
real_t	EstimateLargestEigenvalue (Operator &opr, Vector &v0, int numSteps=10, real_t tolerance=1e-8, int seed=12345)
	Returns an estimate of the largest eigenvalue of the operator opr using the iterative power method.

Detailed Description

PowerMethod helper class to estimate the largest eigenvalue of an operator using the iterative power method.

Definition at line 1017 of file operator.hpp.

Constructor & Destructor Documentation

PowerMethod() [1/3]

mfem::PowerMethod::PowerMethod ( )

inline

Definition at line 1027 of file operator.hpp.

PowerMethod() [2/3]

mfem::PowerMethod::PowerMethod ( )

inline

Definition at line 1029 of file operator.hpp.

PowerMethod() [3/3]

mfem::PowerMethod::PowerMethod ( MPI_Comm comm_ )

inline

Definition at line 1033 of file operator.hpp.

Member Function Documentation

EstimateLargestEigenvalue()

real_t mfem::PowerMethod::EstimateLargestEigenvalue	(	Operator &	opr,
		Vector &	v0,
		int	numSteps = 10,
		real_t	tolerance = 1e-8,
		int	seed = 12345 )

Returns an estimate of the largest eigenvalue of the operator opr using the iterative power method.

v0 is being used as the vector for the iterative process and will contain the eigenvector corresponding to the largest eigenvalue after convergence. The maximum number of iterations may set with numSteps, the relative tolerance with tolerance and the seed of the random initialization of v0 with seed. If seed is 0 v0 will not be random-initialized.

Definition at line 781 of file operator.cpp.

---

The documentation for this class was generated from the following files:

• linalg/operator.hpp
• linalg/operator.cpp