File:  [local] / rpl / lapack / lapack / dtpcon.f
Revision 1.15: download - view: text, annotated - select for diffs - revision graph
Sat Jun 17 10:54:06 2017 UTC (6 years, 10 months ago) by bertrand
Branches: MAIN
CVS tags: HEAD
Mise à jour de lapack.

    1: *> \brief \b DTPCON
    2: *
    3: *  =========== DOCUMENTATION ===========
    4: *
    5: * Online html documentation available at
    6: *            http://www.netlib.org/lapack/explore-html/
    7: *
    8: *> \htmlonly
    9: *> Download DTPCON + dependencies
   10: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dtpcon.f">
   11: *> [TGZ]</a>
   12: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dtpcon.f">
   13: *> [ZIP]</a>
   14: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dtpcon.f">
   15: *> [TXT]</a>
   16: *> \endhtmlonly
   17: *
   18: *  Definition:
   19: *  ===========
   20: *
   21: *       SUBROUTINE DTPCON( NORM, UPLO, DIAG, N, AP, RCOND, WORK, IWORK,
   22: *                          INFO )
   23: *
   24: *       .. Scalar Arguments ..
   25: *       CHARACTER          DIAG, NORM, UPLO
   26: *       INTEGER            INFO, N
   27: *       DOUBLE PRECISION   RCOND
   28: *       ..
   29: *       .. Array Arguments ..
   30: *       INTEGER            IWORK( * )
   31: *       DOUBLE PRECISION   AP( * ), WORK( * )
   32: *       ..
   33: *
   34: *
   35: *> \par Purpose:
   36: *  =============
   37: *>
   38: *> \verbatim
   39: *>
   40: *> DTPCON estimates the reciprocal of the condition number of a packed
   41: *> triangular matrix A, in either the 1-norm or the infinity-norm.
   42: *>
   43: *> The norm of A is computed and an estimate is obtained for
   44: *> norm(inv(A)), then the reciprocal of the condition number is
   45: *> computed as
   46: *>    RCOND = 1 / ( norm(A) * norm(inv(A)) ).
   47: *> \endverbatim
   48: *
   49: *  Arguments:
   50: *  ==========
   51: *
   52: *> \param[in] NORM
   53: *> \verbatim
   54: *>          NORM is CHARACTER*1
   55: *>          Specifies whether the 1-norm condition number or the
   56: *>          infinity-norm condition number is required:
   57: *>          = '1' or 'O':  1-norm;
   58: *>          = 'I':         Infinity-norm.
   59: *> \endverbatim
   60: *>
   61: *> \param[in] UPLO
   62: *> \verbatim
   63: *>          UPLO is CHARACTER*1
   64: *>          = 'U':  A is upper triangular;
   65: *>          = 'L':  A is lower triangular.
   66: *> \endverbatim
   67: *>
   68: *> \param[in] DIAG
   69: *> \verbatim
   70: *>          DIAG is CHARACTER*1
   71: *>          = 'N':  A is non-unit triangular;
   72: *>          = 'U':  A is unit triangular.
   73: *> \endverbatim
   74: *>
   75: *> \param[in] N
   76: *> \verbatim
   77: *>          N is INTEGER
   78: *>          The order of the matrix A.  N >= 0.
   79: *> \endverbatim
   80: *>
   81: *> \param[in] AP
   82: *> \verbatim
   83: *>          AP is DOUBLE PRECISION array, dimension (N*(N+1)/2)
   84: *>          The upper or lower triangular matrix A, packed columnwise in
   85: *>          a linear array.  The j-th column of A is stored in the array
   86: *>          AP as follows:
   87: *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
   88: *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
   89: *>          If DIAG = 'U', the diagonal elements of A are not referenced
   90: *>          and are assumed to be 1.
   91: *> \endverbatim
   92: *>
   93: *> \param[out] RCOND
   94: *> \verbatim
   95: *>          RCOND is DOUBLE PRECISION
   96: *>          The reciprocal of the condition number of the matrix A,
   97: *>          computed as RCOND = 1/(norm(A) * norm(inv(A))).
   98: *> \endverbatim
   99: *>
  100: *> \param[out] WORK
  101: *> \verbatim
  102: *>          WORK is DOUBLE PRECISION array, dimension (3*N)
  103: *> \endverbatim
  104: *>
  105: *> \param[out] IWORK
  106: *> \verbatim
  107: *>          IWORK is INTEGER array, dimension (N)
  108: *> \endverbatim
  109: *>
  110: *> \param[out] INFO
  111: *> \verbatim
  112: *>          INFO is INTEGER
  113: *>          = 0:  successful exit
  114: *>          < 0:  if INFO = -i, the i-th argument had an illegal value
  115: *> \endverbatim
  116: *
  117: *  Authors:
  118: *  ========
  119: *
  120: *> \author Univ. of Tennessee
  121: *> \author Univ. of California Berkeley
  122: *> \author Univ. of Colorado Denver
  123: *> \author NAG Ltd.
  124: *
  125: *> \date December 2016
  126: *
  127: *> \ingroup doubleOTHERcomputational
  128: *
  129: *  =====================================================================
  130:       SUBROUTINE DTPCON( NORM, UPLO, DIAG, N, AP, RCOND, WORK, IWORK,
  131:      $                   INFO )
  132: *
  133: *  -- LAPACK computational routine (version 3.7.0) --
  134: *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
  135: *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
  136: *     December 2016
  137: *
  138: *     .. Scalar Arguments ..
  139:       CHARACTER          DIAG, NORM, UPLO
  140:       INTEGER            INFO, N
  141:       DOUBLE PRECISION   RCOND
  142: *     ..
  143: *     .. Array Arguments ..
  144:       INTEGER            IWORK( * )
  145:       DOUBLE PRECISION   AP( * ), WORK( * )
  146: *     ..
  147: *
  148: *  =====================================================================
  149: *
  150: *     .. Parameters ..
  151:       DOUBLE PRECISION   ONE, ZERO
  152:       PARAMETER          ( ONE = 1.0D+0, ZERO = 0.0D+0 )
  153: *     ..
  154: *     .. Local Scalars ..
  155:       LOGICAL            NOUNIT, ONENRM, UPPER
  156:       CHARACTER          NORMIN
  157:       INTEGER            IX, KASE, KASE1
  158:       DOUBLE PRECISION   AINVNM, ANORM, SCALE, SMLNUM, XNORM
  159: *     ..
  160: *     .. Local Arrays ..
  161:       INTEGER            ISAVE( 3 )
  162: *     ..
  163: *     .. External Functions ..
  164:       LOGICAL            LSAME
  165:       INTEGER            IDAMAX
  166:       DOUBLE PRECISION   DLAMCH, DLANTP
  167:       EXTERNAL           LSAME, IDAMAX, DLAMCH, DLANTP
  168: *     ..
  169: *     .. External Subroutines ..
  170:       EXTERNAL           DLACN2, DLATPS, DRSCL, XERBLA
  171: *     ..
  172: *     .. Intrinsic Functions ..
  173:       INTRINSIC          ABS, DBLE, MAX
  174: *     ..
  175: *     .. Executable Statements ..
  176: *
  177: *     Test the input parameters.
  178: *
  179:       INFO = 0
  180:       UPPER = LSAME( UPLO, 'U' )
  181:       ONENRM = NORM.EQ.'1' .OR. LSAME( NORM, 'O' )
  182:       NOUNIT = LSAME( DIAG, 'N' )
  183: *
  184:       IF( .NOT.ONENRM .AND. .NOT.LSAME( NORM, 'I' ) ) THEN
  185:          INFO = -1
  186:       ELSE IF( .NOT.UPPER .AND. .NOT.LSAME( UPLO, 'L' ) ) THEN
  187:          INFO = -2
  188:       ELSE IF( .NOT.NOUNIT .AND. .NOT.LSAME( DIAG, 'U' ) ) THEN
  189:          INFO = -3
  190:       ELSE IF( N.LT.0 ) THEN
  191:          INFO = -4
  192:       END IF
  193:       IF( INFO.NE.0 ) THEN
  194:          CALL XERBLA( 'DTPCON', -INFO )
  195:          RETURN
  196:       END IF
  197: *
  198: *     Quick return if possible
  199: *
  200:       IF( N.EQ.0 ) THEN
  201:          RCOND = ONE
  202:          RETURN
  203:       END IF
  204: *
  205:       RCOND = ZERO
  206:       SMLNUM = DLAMCH( 'Safe minimum' )*DBLE( MAX( 1, N ) )
  207: *
  208: *     Compute the norm of the triangular matrix A.
  209: *
  210:       ANORM = DLANTP( NORM, UPLO, DIAG, N, AP, WORK )
  211: *
  212: *     Continue only if ANORM > 0.
  213: *
  214:       IF( ANORM.GT.ZERO ) THEN
  215: *
  216: *        Estimate the norm of the inverse of A.
  217: *
  218:          AINVNM = ZERO
  219:          NORMIN = 'N'
  220:          IF( ONENRM ) THEN
  221:             KASE1 = 1
  222:          ELSE
  223:             KASE1 = 2
  224:          END IF
  225:          KASE = 0
  226:    10    CONTINUE
  227:          CALL DLACN2( N, WORK( N+1 ), WORK, IWORK, AINVNM, KASE, ISAVE )
  228:          IF( KASE.NE.0 ) THEN
  229:             IF( KASE.EQ.KASE1 ) THEN
  230: *
  231: *              Multiply by inv(A).
  232: *
  233:                CALL DLATPS( UPLO, 'No transpose', DIAG, NORMIN, N, AP,
  234:      $                      WORK, SCALE, WORK( 2*N+1 ), INFO )
  235:             ELSE
  236: *
  237: *              Multiply by inv(A**T).
  238: *
  239:                CALL DLATPS( UPLO, 'Transpose', DIAG, NORMIN, N, AP,
  240:      $                      WORK, SCALE, WORK( 2*N+1 ), INFO )
  241:             END IF
  242:             NORMIN = 'Y'
  243: *
  244: *           Multiply by 1/SCALE if doing so will not cause overflow.
  245: *
  246:             IF( SCALE.NE.ONE ) THEN
  247:                IX = IDAMAX( N, WORK, 1 )
  248:                XNORM = ABS( WORK( IX ) )
  249:                IF( SCALE.LT.XNORM*SMLNUM .OR. SCALE.EQ.ZERO )
  250:      $            GO TO 20
  251:                CALL DRSCL( N, SCALE, WORK, 1 )
  252:             END IF
  253:             GO TO 10
  254:          END IF
  255: *
  256: *        Compute the estimate of the reciprocal condition number.
  257: *
  258:          IF( AINVNM.NE.ZERO )
  259:      $      RCOND = ( ONE / ANORM ) / AINVNM
  260:       END IF
  261: *
  262:    20 CONTINUE
  263:       RETURN
  264: *
  265: *     End of DTPCON
  266: *
  267:       END

CVSweb interface <joel.bertrand@systella.fr>