File:  [local] / rpl / lapack / lapack / dlaqgb.f
Revision 1.2: download - view: text, annotated - select for diffs - revision graph
Wed Apr 21 13:45:18 2010 UTC (14 years ago) by bertrand
Branches: MAIN
CVS tags: rpl-4_0_17, rpl-4_0_16, rpl-4_0_15, HEAD
En route pour la 4.0.15 !

    1:       SUBROUTINE DLAQGB( M, N, KL, KU, AB, LDAB, R, C, ROWCND, COLCND,
    2:      $                   AMAX, EQUED )
    3: *
    4: *  -- LAPACK auxiliary routine (version 3.2) --
    5: *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
    6: *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
    7: *     November 2006
    8: *
    9: *     .. Scalar Arguments ..
   10:       CHARACTER          EQUED
   11:       INTEGER            KL, KU, LDAB, M, N
   12:       DOUBLE PRECISION   AMAX, COLCND, ROWCND
   13: *     ..
   14: *     .. Array Arguments ..
   15:       DOUBLE PRECISION   AB( LDAB, * ), C( * ), R( * )
   16: *     ..
   17: *
   18: *  Purpose
   19: *  =======
   20: *
   21: *  DLAQGB equilibrates a general M by N band matrix A with KL
   22: *  subdiagonals and KU superdiagonals using the row and scaling factors
   23: *  in the vectors R and C.
   24: *
   25: *  Arguments
   26: *  =========
   27: *
   28: *  M       (input) INTEGER
   29: *          The number of rows of the matrix A.  M >= 0.
   30: *
   31: *  N       (input) INTEGER
   32: *          The number of columns of the matrix A.  N >= 0.
   33: *
   34: *  KL      (input) INTEGER
   35: *          The number of subdiagonals within the band of A.  KL >= 0.
   36: *
   37: *  KU      (input) INTEGER
   38: *          The number of superdiagonals within the band of A.  KU >= 0.
   39: *
   40: *  AB      (input/output) DOUBLE PRECISION array, dimension (LDAB,N)
   41: *          On entry, the matrix A in band storage, in rows 1 to KL+KU+1.
   42: *          The j-th column of A is stored in the j-th column of the
   43: *          array AB as follows:
   44: *          AB(ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
   45: *
   46: *          On exit, the equilibrated matrix, in the same storage format
   47: *          as A.  See EQUED for the form of the equilibrated matrix.
   48: *
   49: *  LDAB    (input) INTEGER
   50: *          The leading dimension of the array AB.  LDA >= KL+KU+1.
   51: *
   52: *  R       (input) DOUBLE PRECISION array, dimension (M)
   53: *          The row scale factors for A.
   54: *
   55: *  C       (input) DOUBLE PRECISION array, dimension (N)
   56: *          The column scale factors for A.
   57: *
   58: *  ROWCND  (input) DOUBLE PRECISION
   59: *          Ratio of the smallest R(i) to the largest R(i).
   60: *
   61: *  COLCND  (input) DOUBLE PRECISION
   62: *          Ratio of the smallest C(i) to the largest C(i).
   63: *
   64: *  AMAX    (input) DOUBLE PRECISION
   65: *          Absolute value of largest matrix entry.
   66: *
   67: *  EQUED   (output) CHARACTER*1
   68: *          Specifies the form of equilibration that was done.
   69: *          = 'N':  No equilibration
   70: *          = 'R':  Row equilibration, i.e., A has been premultiplied by
   71: *                  diag(R).
   72: *          = 'C':  Column equilibration, i.e., A has been postmultiplied
   73: *                  by diag(C).
   74: *          = 'B':  Both row and column equilibration, i.e., A has been
   75: *                  replaced by diag(R) * A * diag(C).
   76: *
   77: *  Internal Parameters
   78: *  ===================
   79: *
   80: *  THRESH is a threshold value used to decide if row or column scaling
   81: *  should be done based on the ratio of the row or column scaling
   82: *  factors.  If ROWCND < THRESH, row scaling is done, and if
   83: *  COLCND < THRESH, column scaling is done.
   84: *
   85: *  LARGE and SMALL are threshold values used to decide if row scaling
   86: *  should be done based on the absolute size of the largest matrix
   87: *  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
   88: *
   89: *  =====================================================================
   90: *
   91: *     .. Parameters ..
   92:       DOUBLE PRECISION   ONE, THRESH
   93:       PARAMETER          ( ONE = 1.0D+0, THRESH = 0.1D+0 )
   94: *     ..
   95: *     .. Local Scalars ..
   96:       INTEGER            I, J
   97:       DOUBLE PRECISION   CJ, LARGE, SMALL
   98: *     ..
   99: *     .. External Functions ..
  100:       DOUBLE PRECISION   DLAMCH
  101:       EXTERNAL           DLAMCH
  102: *     ..
  103: *     .. Intrinsic Functions ..
  104:       INTRINSIC          MAX, MIN
  105: *     ..
  106: *     .. Executable Statements ..
  107: *
  108: *     Quick return if possible
  109: *
  110:       IF( M.LE.0 .OR. N.LE.0 ) THEN
  111:          EQUED = 'N'
  112:          RETURN
  113:       END IF
  114: *
  115: *     Initialize LARGE and SMALL.
  116: *
  117:       SMALL = DLAMCH( 'Safe minimum' ) / DLAMCH( 'Precision' )
  118:       LARGE = ONE / SMALL
  119: *
  120:       IF( ROWCND.GE.THRESH .AND. AMAX.GE.SMALL .AND. AMAX.LE.LARGE )
  121:      $     THEN
  122: *
  123: *        No row scaling
  124: *
  125:          IF( COLCND.GE.THRESH ) THEN
  126: *
  127: *           No column scaling
  128: *
  129:             EQUED = 'N'
  130:          ELSE
  131: *
  132: *           Column scaling
  133: *
  134:             DO 20 J = 1, N
  135:                CJ = C( J )
  136:                DO 10 I = MAX( 1, J-KU ), MIN( M, J+KL )
  137:                   AB( KU+1+I-J, J ) = CJ*AB( KU+1+I-J, J )
  138:    10          CONTINUE
  139:    20       CONTINUE
  140:             EQUED = 'C'
  141:          END IF
  142:       ELSE IF( COLCND.GE.THRESH ) THEN
  143: *
  144: *        Row scaling, no column scaling
  145: *
  146:          DO 40 J = 1, N
  147:             DO 30 I = MAX( 1, J-KU ), MIN( M, J+KL )
  148:                AB( KU+1+I-J, J ) = R( I )*AB( KU+1+I-J, J )
  149:    30       CONTINUE
  150:    40    CONTINUE
  151:          EQUED = 'R'
  152:       ELSE
  153: *
  154: *        Row and column scaling
  155: *
  156:          DO 60 J = 1, N
  157:             CJ = C( J )
  158:             DO 50 I = MAX( 1, J-KU ), MIN( M, J+KL )
  159:                AB( KU+1+I-J, J ) = CJ*R( I )*AB( KU+1+I-J, J )
  160:    50       CONTINUE
  161:    60    CONTINUE
  162:          EQUED = 'B'
  163:       END IF
  164: *
  165:       RETURN
  166: *
  167: *     End of DLAQGB
  168: *
  169:       END

CVSweb interface <joel.bertrand@systella.fr>