cp_dbcsr_diag Module Reference

Interface to (sca)lapack for the Cholesky based procedures. More...

Functions/Subroutines
subroutine, public	cp_dbcsr_syevd (matrix, eigenvectors, eigenvalues, para_env, blacs_env)
	...
subroutine, public	cp_dbcsr_syevx (matrix, eigenvectors, eigenvalues, neig, work_syevx, para_env, blacs_env)
	compute eigenvalues and optionally eigenvectors of a real symmetric matrix using scalapack. If eigenvectors are required this routine will replicate a full matrix on each CPU... if more than a handful of vectors are needed, use cp_dbcsr_syevd instead
subroutine, public	cp_dbcsr_heevd (matrix_re, matrix_im, eigenvectors_re, eigenvectors_im, eigenvalues, para_env, blacs_env)
	...
subroutine, public	cp_dbcsr_power (matrix, exponent, threshold, n_dependent, para_env, blacs_env, verbose, eigenvectors, eigenvalues)
	...

Detailed Description

Interface to (sca)lapack for the Cholesky based procedures.

Author

VW

Date

2009-11-09

Version

0.8

Modification history:

• Created 2009-11-09

Function/Subroutine Documentation

cp_dbcsr_syevd()

subroutine, public cp_dbcsr_diag::cp_dbcsr_syevd	(	type(dbcsr_type)	matrix,
		type(dbcsr_type)	eigenvectors,
		real(kind=dp), dimension(:), intent(out)	eigenvalues,
		type(mp_para_env_type), pointer	para_env,
		type(cp_blacs_env_type), pointer	blacs_env )

...

Parameters

matrix	...
eigenvectors	...
eigenvalues	...
para_env	...
blacs_env	...

Definition at line 66 of file cp_dbcsr_diag.F.

Here is the call graph for this function:

Here is the caller graph for this function:

cp_dbcsr_syevx()

subroutine, public cp_dbcsr_diag::cp_dbcsr_syevx	(	type(dbcsr_type), pointer	matrix,
		type(dbcsr_type), optional, pointer	eigenvectors,
		real(kind=dp), dimension(:), intent(out)	eigenvalues,
		integer, intent(in), optional	neig,
		real(kind=dp), intent(in), optional	work_syevx,
		type(mp_para_env_type), pointer	para_env,
		type(cp_blacs_env_type), pointer	blacs_env )

compute eigenvalues and optionally eigenvectors of a real symmetric matrix using scalapack. If eigenvectors are required this routine will replicate a full matrix on each CPU... if more than a handful of vectors are needed, use cp_dbcsr_syevd instead

Parameters

matrix	...
eigenvectors	...
eigenvalues	...
neig	...
work_syevx	...
para_env	...
blacs_env	...

matrix is supposed to be in upper triangular form, and overwritten by this routine

neig is the number of vectors needed (default all) work_syevx evec calculation only, is the fraction of the working buffer allowed (1.0 use full buffer) reducing this saves time, but might cause the routine to fail

Definition at line 124 of file cp_dbcsr_diag.F.

Here is the call graph for this function:

Here is the caller graph for this function:

cp_dbcsr_heevd()

subroutine, public cp_dbcsr_diag::cp_dbcsr_heevd	(	type(dbcsr_type), optional	matrix_re,
		type(dbcsr_type), optional	matrix_im,
		type(dbcsr_type), optional	eigenvectors_re,
		type(dbcsr_type), optional	eigenvectors_im,
		real(kind=dp), dimension(:), intent(out)	eigenvalues,
		type(mp_para_env_type), pointer	para_env,
		type(cp_blacs_env_type), pointer	blacs_env )

...

Parameters

matrix_re	...
matrix_im	...
eigenvectors_re	...
eigenvectors_im	...
eigenvalues	...
para_env	...
blacs_env	...

Definition at line 184 of file cp_dbcsr_diag.F.

Here is the call graph for this function:

Here is the caller graph for this function:

cp_dbcsr_power()

subroutine, public cp_dbcsr_diag::cp_dbcsr_power	(	type(dbcsr_type), intent(inout)	matrix,
		real(dp), intent(in)	exponent,
		real(dp), intent(in)	threshold,
		integer, intent(out)	n_dependent,
		type(mp_para_env_type), pointer	para_env,
		type(cp_blacs_env_type), pointer	blacs_env,
		logical, intent(in), optional	verbose,
		type(dbcsr_type), intent(inout), optional	eigenvectors,
		real(kind=dp), dimension(2), intent(out), optional	eigenvalues )

...

Parameters

matrix	...
exponent	...
threshold	...
n_dependent	...
para_env	...
blacs_env	...
verbose	...
eigenvectors	...
eigenvalues	...

Definition at line 278 of file cp_dbcsr_diag.F.

Here is the call graph for this function:

Here is the caller graph for this function: