hegv_2stage#

Functions

void chegv_2stage(
    const INT            itype,
    const char*          jobz,
    const char*          uplo,
    const INT            n,
          c64*  restrict A,
    const INT            lda,
          c64*  restrict B,
    const INT            ldb,
          f32*  restrict W,
          c64*  restrict work,
    const INT            lwork,
          f32*  restrict rwork,
          INT*           info
);
void chegv_2stage(const INT itype, const char *jobz, const char *uplo, const INT n, c64 *restrict A, const INT lda, c64 *restrict B, const INT ldb, f32 *restrict W, c64 *restrict work, const INT lwork, f32 *restrict rwork, INT *info)#

CHEGV_2STAGE computes all the eigenvalues, and optionally, the eigenvectors of a complex generalized Hermitian-definite eigenproblem, of the form A*x=(lambda)*B*x, A*B*x=(lambda)*x, or B*A*x=(lambda)*x.

Here A and B are assumed to be Hermitian and B is also positive definite. This routine uses the 2-stage technique for the reduction to tridiagonal which showed higher performance on recent architecture and for large sizes N>2000.

Parameters

in
itype

= 1: A*x = lambda*B*x; = 2: A*B*x = lambda*x; = 3: B*A*x = lambda*x

in
jobz

= ‘N’: eigenvalues only; = ‘V’: not available in this release.

in
uplo

= ‘U’: upper triangles stored; = ‘L’: lower triangles stored

in
n

The order of the matrices A and B. n >= 0.

inout
A

Hermitian matrix A. On exit, eigenvectors if jobz=’V’.

in
lda

Leading dimension of A. lda >= max(1, n).

inout
B

Hermitian positive definite B. On exit, Cholesky factor.

in
ldb

Leading dimension of B. ldb >= max(1, n).

out
W

Eigenvalues in ascending order.

out
work

Workspace. On exit, work[0] = optimal LWORK.

in
lwork

Length of work. If -1, workspace query.

out
rwork

Real workspace, dimension max(1, 3*n-2).

out
info

  • = 0: success; < 0: illegal argument; > 0: CPOTRF/CHEEV error.

Functions

void zhegv_2stage(
    const INT            itype,
    const char*          jobz,
    const char*          uplo,
    const INT            n,
          c128* restrict A,
    const INT            lda,
          c128* restrict B,
    const INT            ldb,
          f64*  restrict W,
          c128* restrict work,
    const INT            lwork,
          f64*  restrict rwork,
          INT*           info
);
void zhegv_2stage(const INT itype, const char *jobz, const char *uplo, const INT n, c128 *restrict A, const INT lda, c128 *restrict B, const INT ldb, f64 *restrict W, c128 *restrict work, const INT lwork, f64 *restrict rwork, INT *info)#

ZHEGV_2STAGE computes all the eigenvalues, and optionally, the eigenvectors of a complex generalized Hermitian-definite eigenproblem, of the form A*x=(lambda)*B*x, A*B*x=(lambda)*x, or B*A*x=(lambda)*x.

Here A and B are assumed to be Hermitian and B is also positive definite. This routine uses the 2-stage technique for the reduction to tridiagonal which showed higher performance on recent architecture and for large sizes N>2000.

Parameters

in
itype

= 1: A*x = lambda*B*x; = 2: A*B*x = lambda*x; = 3: B*A*x = lambda*x

in
jobz

= ‘N’: eigenvalues only; = ‘V’: not available in this release.

in
uplo

= ‘U’: upper triangles stored; = ‘L’: lower triangles stored

in
n

The order of the matrices A and B. n >= 0.

inout
A

Hermitian matrix A. On exit, eigenvectors if jobz=’V’.

in
lda

Leading dimension of A. lda >= max(1, n).

inout
B

Hermitian positive definite B. On exit, Cholesky factor.

in
ldb

Leading dimension of B. ldb >= max(1, n).

out
W

Eigenvalues in ascending order.

out
work

Workspace. On exit, work[0] = optimal LWORK.

in
lwork

Length of work. If -1, workspace query.

out
rwork

Real workspace, dimension max(1, 3*n-2).

out
info

  • = 0: success; < 0: illegal argument; > 0: ZPOTRF/ZHEEV error.

On this page

This Page