Driver

\(Ax = \lambda B x\), etc. More...

Functions
template<typename scalar_t >
void	slate::hegv (int64_t itype, HermitianMatrix< scalar_t > &A, HermitianMatrix< scalar_t > &B, std::vector< blas::real_type< scalar_t > > &Lambda, Matrix< scalar_t > &Z, Options const &opts)
	Distributed parallel computation of all the eigenvalues, and optionally, the eigenvectors of a complex generalized Hermitian-definite eigenproblem, of the form:

Detailed Description

\(Ax = \lambda B x\), etc.

Function Documentation

hegv()

template<typename scalar_t >

void slate::hegv	(	int64_t	itype,
		HermitianMatrix< scalar_t > &	A,
		HermitianMatrix< scalar_t > &	B,
		std::vector< blas::real_type< scalar_t > > &	Lambda,
		Matrix< scalar_t > &	Z,
		Options const &	opts
	)

Distributed parallel computation of all the eigenvalues, and optionally, the eigenvectors of a complex generalized Hermitian-definite eigenproblem, of the form:

itype	Problem
itype = 1	\(A z = \lambda B z\)
itype = 2	\(A B z = \lambda z\)
itype = 3	\(B A z = \lambda z\)

Here A and B are assumed to be Hermitian and B is also positive definite.

Template Parameters

scalar_t

One of float, double, std::complex<float>, std::complex<double>.

Parameters

[in]	itype	itype = 1: Compute \(A z = \lambda B z\); itype = 2: Compute \(A B z = \lambda z\); itype = 3: Compute \(B A z = \lambda z\).
[in,out]	A	On entry, the n-by-n Hermitian matrix \(A\). On exit, contents are destroyed.
[in,out]	B	On entry, the n-by-n Hermitian positive definite matrix \(B\). On exit, B is overwritten by the triangular factor U or L from the Cholesky factorization \(B = U^H U\) or \(B = L L^H\).
[out]	Lambda	The vector Lambda of length n. If successful, the eigenvalues in ascending order.
[out]	Z	On entry, if Z is empty, does not compute eigenvectors. On exit, orthonormal eigenvectors of the matrix A.
[in]	opts	Additional options, as map of name = value pairs. Possible options: Option::Lookahead: Number of panels to overlap with matrix updates. lookahead >= 0. Default 1. Option::InnerBlocking: Inner blocking to use for panel. Default 16. Option::MaxPanelThreads: Number of threads to use for panel. Default omp_get_max_threads()/2. Option::Target: Implementation to target. Possible values: HostTask: OpenMP tasks on CPU host [default]. HostNest: nested OpenMP parallel for loop on CPU host. HostBatch: batched BLAS on CPU host. Devices: batched BLAS on GPU device.

TODO: return value