Auxilliary Routines Parameters APIs#
LAMCH#
- group lamch
Functions
-
template<typename T>
T lamch(char *cmach)# LAMCH determines single precision machine parameters.
Purpose:
LAMCH determines single precision machine parameters.
- Parameters:
CMACH – [in]
CMACH is CHARACTER*1
Specifies the value to be returned by SLAMCH:
= ‘E’ or ‘e’, SLAMCH := eps
= ‘S’ or ‘s , SLAMCH := sfmin
= ‘B’ or ‘b’, SLAMCH := base
= ‘P’ or ‘p’, SLAMCH := eps*base
= ‘N’ or ‘n’, SLAMCH := t
= ‘R’ or ‘r’, SLAMCH := rnd
= ‘M’ or ‘m’, SLAMCH := emin
= ‘U’ or ‘u’, SLAMCH := rmin
= ‘L’ or ‘l’, SLAMCH := emax
= ‘O’ or ‘o’, SLAMCH := rmax
where
eps = relative machine precision
sfmin = safe minimum, such that 1/sfmin does not overflow
base = base of the machine
prec = eps*base
t = number of (base) digits in the mantissa
rnd = 1.0 when rounding occurs in addition, 0.0 otherwise
emin = minimum exponent before (gradual) underflow
rmin = underflow threshold - base**(emin-1)
emax = largest exponent before overflow
rmax = overflow threshold - (base**emax)*(1-eps)- Returns:
T Template based return value of the function.
-
template<typename T>
LABAD#
- group labad
Functions
-
template<typename T>
void labad(T *small_val, T *large)# LABAD takes as input the values computed by LAMCH for underflow and
overflow, and returns the square root of each of these values.
Purpose:
LABAD takes as input the values computed by SLAMCH for underflow and overflow, and returns the square root of each of these values if the log of LARGE is sufficiently large. This subroutine is intended to identify machines with a large exponent range, such as the Crays, and redefine the underflow and overflow limits to be the square roots of the values computed by SLAMCH. This subroutine is needed because SLAMCH does not compensate for poor arithmetic in the upper half of the exponent range, as is found on a Cray.
- Parameters:
SMALL_VAL – [inout]
SMALL_VAL is REAL
On entry, the underflow threshold as computed by SLAMCH. On exit, if LOG10(LARGE) is sufficiently large, the square root of SMALL, otherwise unchanged.
LARGE – [inout]
LARGE is REAL
On entry, the overflow threshold as computed by SLAMCH. On exit, if LOG10(LARGE) is sufficiently large, the square root of LARGE, otherwise unchanged.
-
template<typename T>
ILAVER#
- group ilaver
Functions
-
inline void ilaver(integer *vers_major, integer *vers_minor, integer *vers_patch__)#
ILAVER returns the LAPACK version.
Purpose:
This subroutine returns the LAPACK version.
- Parameters:
VERS_MAJOR – [out]
VERS_MAJOR is INTEGER
return the lapack major version
VERS_MINOR – [out]
VERS_MINOR is INTEGER
return the lapack minor version from the major version
VERS_PATCH – [out]
VERS_PATCH is INTEGER
return the lapack patch version from the minor version
-
inline void ilaver(integer *vers_major, integer *vers_minor, integer *vers_patch__)#
IPARAM2STAGE#
- group iparam2stage
Functions
-
inline integer iparam2stage(integer *ispec, char *name, char *opts, integer *ni, integer *nbi, integer *ibi, integer *nxi)#
IPARAM2STAGE program sets problem and machine dependent parameters.
Purpose:
This program sets problem and machine dependent parameters useful for xHETRD_2STAGE, xHETRD_HE2HB, xHETRD_HB2ST, xGEBRD_2STAGE, xGEBRD_GE2GB, xGEBRD_GB2BD and related subroutines for eigenvalue problems. It is called whenever ILAENV is called with 17 <= ISPEC <= 21. It is called whenever ILAENV2STAGE is called with 1 <= ISPEC <= 5 with a direct conversion ISPEC + 16.
- Parameters:
ISPEC – [in]
ISPEC is integer scalar
ISPEC specifies which tunable parameter IPARAM2STAGE should return.
ISPEC=17: the optimal blocksize nb for the reduction to BAND
ISPEC=18: the optimal blocksize ib for the eigenvectors singular vectors update routine
ISPEC=19: The length of the array that store the Housholder representation for the second stage Band to Tridiagonal or Bidiagonal
ISPEC=20: The workspace needed for the routine in input.
ISPEC=21: For future release.NAME – [in]
NAME is character string
Name of the calling subroutine
OPTS – [in]
OPTS is CHARACTER*(*)
The character options to the subroutine NAME, concatenated into a single character string. For example, UPLO = ‘U’, TRANS = ‘T’, and DIAG = ‘N’ for a triangular routine would be specified as OPTS = ‘UTN’.
NI – [in] NI is INTEGER which is the size of the matrix
NBI – [in]
NBI is INTEGER which is the used in the reduciton,
(e.g., the size of the band), needed to compute workspace and LHOUS2.
IBI – [in] IBI is INTEGER which represent the IB of the reduciton, needed to compute workspace and LHOUS2.
NXI – [in] NXI is INTEGER needed in the future release.
- Returns:
INTEGER Return value of the function.
-
inline integer iparam2stage(integer *ispec, char *name, char *opts, integer *ni, integer *nbi, integer *ibi, integer *nxi)#
ILADIAG#
- group iladiag
Functions
-
inline integer iladiag(char *diag)#
ILADIAG translated from a character string specifying if a matrix
has unit diagonal or not to the relevant BLAST-specified integer constant.
Purpose:
This subroutine translated from a character string specifying if a matrix has unit diagonal or not to the relevant BLAST-specified integer constant. ILADIAG returns an INTEGER. If ILADIAG < 0, then the input is not a character indicating a unit or non-unit diagonal. Otherwise ILADIAG returns the constant value corresponding to DIAG.
- Parameters:
diag – [in]
DIAG is character pointer.
= N means non-unit diagonal.
= U means unit diagonal.- Returns:
INTEGER Return value of the function.
-
inline integer iladiag(char *diag)#
ILATRANS#
- group ilatrans
Functions
-
inline integer ilatrans(char *trans)#
ILATRANS translates from a character string specifying a transposition
operation to the relevant BLAST-specified integer constant.
Purpose:
This subroutine translates from a character string specifying a transposition operation to the relevant BLAST-specified integer constant. ILATRANS returns an INTEGER. If ILATRANS < 0, then the input is not a character indicating a transposition operator. Otherwise ILATRANS returns the constant value corresponding to TRANS
- Parameters:
trans – [in]
trans is charater pointer.
= N means no transposition
= T means transposition
= C means conjugate transposition
- Returns:
INTEGER Return value of the function.
-
inline integer ilatrans(char *trans)#
ILAPREC#
- group ilaprec
Functions
-
inline integer ilaprec(char *prec)#
ILAPREC translated from a character string specifying an intermediate
precision to the relevant BLAST-specified integer constant.
Purpose:
This subroutine translated from a character string specifying an intermediate precision to the relevant BLAST-specified integer constant. ILAPREC returns an INTEGER. If ILAPREC < 0, then the input is not a character indicating a supported intermediate precision. Otherwise ILAPREC returns the constant value corresponding to PREC.
- Parameters:
prec – [in]
PREC is character pointer.
= S means single precision
= D means double precision
= I means indigenous precision
= X or E means extra precision.
Other values returns invalid result.
- Returns:
INTEGER Return value of the function.
-
inline integer ilaprec(char *prec)#
ILAENV2STAGE#
- group ilaenv2stage
Functions
-
inline integer ilaenv2stage(integer *ispec, char *name, char *opts, integer *n1, integer *n2, integer *n3, integer *n4)#
ILAENV2STAGE is called from the LAPACK routines to choose problem-dependent
parameters for the local environment.
Purpose:
ILAENV2STAGE is called from the LAPACK routines to choose problem-dependent parameters for the local environment. See ISPEC for a description of the parameters. It sets problem and machine dependent parameters useful for *_2STAGE and related subroutines. ILAENV2STAGE returns an INTEGER if ILAENV2STAGE >= 0: ILAENV2STAGE returns the value of the parameter specified by ISPEC if ILAENV2STAGE < 0: if ILAENV2STAGE = -k, the k-th argument had an illegal value. This version provides a set of parameters which should give good, but not optimal, performance on many of the currently available computers for the 2-stage solvers. Users are encouraged to modify this subroutine to set the tuning parameters for their particular machine using the option and problem size information in the arguments. This routine will not function correctly if it is converted to all lower case. Converting it to all upper case is allowed.
- Parameters:
ISPEC – [in]
ISPEC is INTEGER
Specifies the parameter to be returned as the value of ILAENV2STAGE.
= 1: the optimal blocksize nb for the reduction to BAND
= 2: the optimal blocksize ib for the eigenvectors singular vectors update routine
= 3: The length of the array that store the Housholder representation for the second stage Band to Tridiagonal or Bidiagonal
= 4: The workspace needed for the routine in input.
= 5: For future release.NAME – [in]
NAME is CHARACTER*(*)
The name of the calling subroutine, in either upper case or lower case.
OPTS – [in]
OPTS is CHARACTER*(*)
The character options to the subroutine NAME, concatenated into a single character string. For example, UPLO = ‘U’, TRANS = ‘T’, and DIAG = ‘N’ for a triangular routine would be specified as OPTS = ‘UTN’.
N1 – [in] N1 is INTEGER
N2 – [in] N2 is INTEGER
N3 – [in] N3 is INTEGER
N4 – [in]
N4 is INTEGER
Problem dimensions for the subroutine NAME; these may not all be required.
- Returns:
INTEGER Return value of the function.
-
inline integer ilaenv2stage(integer *ispec, char *name, char *opts, integer *n1, integer *n2, integer *n3, integer *n4)#
ILAUPLO#
- group ilauplo
Functions
-
inline integer ilauplo(char *uplo)#
ILAUPLO translated from a character string specifying a upper- or
lower-triangular matrix to the relevant BLAST-specified integer constant.
Purpose:
This subroutine translated from a character string specifying a upper- or lower-triangular matrix to the relevant BLAST-specified integer constant. ILAUPLO returns an INTEGER. If ILAUPLO < 0, then the input is not a character indicating an upper- or lower-triangular matrix. Otherwise ILAUPLO returns the constant value corresponding to UPLO
- Parameters:
uplo – [in]
uplo is charater pointer.
= ‘L’ means lower triangular matrix
= ‘U’ means upper triangular matrix
Other values, returns invalid result.
- Returns:
INTEGER Return value of the function.
-
inline integer ilauplo(char *uplo)#
LATRANS#
- group latrans
Functions
-
inline void hla_transtype(char *ret_val, integer *trans)#
CHLA_TRANSTYPE subroutine translates from a BLAST-specified integer constant
to the character string specifying a transposition operation.
Purpose:
This subroutine translates from a BLAST-specified integer constant to the character string specifying a transposition operation. CHLA_TRANSTYPE returns an CHARACTER*1. If CHLA_TRANSTYPE is 'X', then input is not an integer indicating a transposition operator. Otherwise CHLA_TRANSTYPE returns the constant value corresponding to TRANS.
-
inline void hla_transtype(char *ret_val, integer *trans)#