Annotation of rpl/lapack/lapack/zla_porpvgrw.f, revision 1.12

1.8       bertrand    1: *> \brief \b ZLA_PORPVGRW computes the reciprocal pivot growth factor norm(A)/norm(U) for a symmetric or Hermitian positive-definite matrix.
1.5       bertrand    2: *
                      3: *  =========== DOCUMENTATION ===========
                      4: *
                      5: * Online html documentation available at 
                      6: *            http://www.netlib.org/lapack/explore-html/ 
                      7: *
                      8: *> \htmlonly
                      9: *> Download ZLA_PORPVGRW + dependencies 
                     10: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/zla_porpvgrw.f"> 
                     11: *> [TGZ]</a> 
                     12: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/zla_porpvgrw.f"> 
                     13: *> [ZIP]</a> 
                     14: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/zla_porpvgrw.f"> 
                     15: *> [TXT]</a>
                     16: *> \endhtmlonly 
                     17: *
                     18: *  Definition:
                     19: *  ===========
                     20: *
                     21: *       DOUBLE PRECISION FUNCTION ZLA_PORPVGRW( UPLO, NCOLS, A, LDA, AF, 
                     22: *                                               LDAF, WORK )
                     23: * 
                     24: *       .. Scalar Arguments ..
                     25: *       CHARACTER*1        UPLO
                     26: *       INTEGER            NCOLS, LDA, LDAF
                     27: *       ..
                     28: *       .. Array Arguments ..
                     29: *       COMPLEX*16         A( LDA, * ), AF( LDAF, * )
                     30: *       DOUBLE PRECISION   WORK( * )
                     31: *       ..
                     32: *  
                     33: *
                     34: *> \par Purpose:
                     35: *  =============
                     36: *>
                     37: *> \verbatim
                     38: *>
                     39: *> 
                     40: *> ZLA_PORPVGRW computes the reciprocal pivot growth factor
                     41: *> norm(A)/norm(U). The "max absolute element" norm is used. If this is
                     42: *> much less than 1, the stability of the LU factorization of the
                     43: *> (equilibrated) matrix A could be poor. This also means that the
                     44: *> solution X, estimated condition numbers, and error bounds could be
                     45: *> unreliable.
                     46: *> \endverbatim
                     47: *
                     48: *  Arguments:
                     49: *  ==========
                     50: *
                     51: *> \param[in] UPLO
                     52: *> \verbatim
                     53: *>          UPLO is CHARACTER*1
                     54: *>       = 'U':  Upper triangle of A is stored;
                     55: *>       = 'L':  Lower triangle of A is stored.
                     56: *> \endverbatim
                     57: *>
                     58: *> \param[in] NCOLS
                     59: *> \verbatim
                     60: *>          NCOLS is INTEGER
                     61: *>     The number of columns of the matrix A. NCOLS >= 0.
                     62: *> \endverbatim
                     63: *>
                     64: *> \param[in] A
                     65: *> \verbatim
                     66: *>          A is COMPLEX*16 array, dimension (LDA,N)
                     67: *>     On entry, the N-by-N matrix A.
                     68: *> \endverbatim
                     69: *>
                     70: *> \param[in] LDA
                     71: *> \verbatim
                     72: *>          LDA is INTEGER
                     73: *>     The leading dimension of the array A.  LDA >= max(1,N).
                     74: *> \endverbatim
                     75: *>
                     76: *> \param[in] AF
                     77: *> \verbatim
                     78: *>          AF is COMPLEX*16 array, dimension (LDAF,N)
                     79: *>     The triangular factor U or L from the Cholesky factorization
                     80: *>     A = U**T*U or A = L*L**T, as computed by ZPOTRF.
                     81: *> \endverbatim
                     82: *>
                     83: *> \param[in] LDAF
                     84: *> \verbatim
                     85: *>          LDAF is INTEGER
                     86: *>     The leading dimension of the array AF.  LDAF >= max(1,N).
                     87: *> \endverbatim
                     88: *>
                     89: *> \param[in] WORK
                     90: *> \verbatim
1.11      bertrand   91: *>          WORK is DOUBLE PRECISION array, dimension (2*N)
1.5       bertrand   92: *> \endverbatim
                     93: *
                     94: *  Authors:
                     95: *  ========
                     96: *
                     97: *> \author Univ. of Tennessee 
                     98: *> \author Univ. of California Berkeley 
                     99: *> \author Univ. of Colorado Denver 
                    100: *> \author NAG Ltd. 
                    101: *
1.11      bertrand  102: *> \date June 2016
1.5       bertrand  103: *
                    104: *> \ingroup complex16POcomputational
                    105: *
                    106: *  =====================================================================
1.1       bertrand  107:       DOUBLE PRECISION FUNCTION ZLA_PORPVGRW( UPLO, NCOLS, A, LDA, AF, 
                    108:      $                                        LDAF, WORK )
                    109: *
1.11      bertrand  110: *  -- LAPACK computational routine (version 3.6.1) --
1.5       bertrand  111: *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
                    112: *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
1.11      bertrand  113: *     June 2016
1.1       bertrand  114: *
                    115: *     .. Scalar Arguments ..
                    116:       CHARACTER*1        UPLO
                    117:       INTEGER            NCOLS, LDA, LDAF
                    118: *     ..
                    119: *     .. Array Arguments ..
                    120:       COMPLEX*16         A( LDA, * ), AF( LDAF, * )
                    121:       DOUBLE PRECISION   WORK( * )
                    122: *     ..
                    123: *
                    124: *  =====================================================================
                    125: *
                    126: *     .. Local Scalars ..
                    127:       INTEGER            I, J
                    128:       DOUBLE PRECISION   AMAX, UMAX, RPVGRW
                    129:       LOGICAL            UPPER
                    130:       COMPLEX*16         ZDUM
                    131: *     ..
                    132: *     .. External Functions ..
                    133:       EXTERNAL           LSAME, ZLASET
                    134:       LOGICAL            LSAME
                    135: *     ..
                    136: *     .. Intrinsic Functions ..
                    137:       INTRINSIC          ABS, MAX, MIN, REAL, DIMAG
                    138: *     ..
                    139: *     .. Statement Functions ..
                    140:       DOUBLE PRECISION   CABS1
                    141: *     ..
                    142: *     .. Statement Function Definitions ..
                    143:       CABS1( ZDUM ) = ABS( DBLE( ZDUM ) ) + ABS( DIMAG( ZDUM ) )
                    144: *     ..
                    145: *     .. Executable Statements ..
                    146:       UPPER = LSAME( 'Upper', UPLO )
                    147: *
                    148: *     DPOTRF will have factored only the NCOLSxNCOLS leading minor, so
                    149: *     we restrict the growth search to that minor and use only the first
                    150: *     2*NCOLS workspace entries.
                    151: *
                    152:       RPVGRW = 1.0D+0
                    153:       DO I = 1, 2*NCOLS
                    154:          WORK( I ) = 0.0D+0
                    155:       END DO
                    156: *
                    157: *     Find the max magnitude entry of each column.
                    158: *
                    159:       IF ( UPPER ) THEN
                    160:          DO J = 1, NCOLS
                    161:             DO I = 1, J
                    162:                WORK( NCOLS+J ) =
                    163:      $              MAX( CABS1( A( I, J ) ), WORK( NCOLS+J ) )
                    164:             END DO
                    165:          END DO
                    166:       ELSE
                    167:          DO J = 1, NCOLS
                    168:             DO I = J, NCOLS
                    169:                WORK( NCOLS+J ) =
                    170:      $              MAX( CABS1( A( I, J ) ), WORK( NCOLS+J ) )
                    171:             END DO
                    172:          END DO
                    173:       END IF
                    174: *
                    175: *     Now find the max magnitude entry of each column of the factor in
                    176: *     AF.  No pivoting, so no permutations.
                    177: *
                    178:       IF ( LSAME( 'Upper', UPLO ) ) THEN
                    179:          DO J = 1, NCOLS
                    180:             DO I = 1, J
                    181:                WORK( J ) = MAX( CABS1( AF( I, J ) ), WORK( J ) )
                    182:             END DO
                    183:          END DO
                    184:       ELSE
                    185:          DO J = 1, NCOLS
                    186:             DO I = J, NCOLS
                    187:                WORK( J ) = MAX( CABS1( AF( I, J ) ), WORK( J ) )
                    188:             END DO
                    189:          END DO
                    190:       END IF
                    191: *
                    192: *     Compute the *inverse* of the max element growth factor.  Dividing
                    193: *     by zero would imply the largest entry of the factor's column is
                    194: *     zero.  Than can happen when either the column of A is zero or
                    195: *     massive pivots made the factor underflow to zero.  Neither counts
                    196: *     as growth in itself, so simply ignore terms with zero
                    197: *     denominators.
                    198: *
                    199:       IF ( LSAME( 'Upper', UPLO ) ) THEN
                    200:          DO I = 1, NCOLS
                    201:             UMAX = WORK( I )
                    202:             AMAX = WORK( NCOLS+I )
                    203:             IF ( UMAX /= 0.0D+0 ) THEN
                    204:                RPVGRW = MIN( AMAX / UMAX, RPVGRW )
                    205:             END IF
                    206:          END DO
                    207:       ELSE
                    208:          DO I = 1, NCOLS
                    209:             UMAX = WORK( I )
                    210:             AMAX = WORK( NCOLS+I )
                    211:             IF ( UMAX /= 0.0D+0 ) THEN
                    212:                RPVGRW = MIN( AMAX / UMAX, RPVGRW )
                    213:             END IF
                    214:          END DO
                    215:       END IF
                    216: 
                    217:       ZLA_PORPVGRW = RPVGRW
                    218:       END

CVSweb interface <joel.bertrand@systella.fr>