1
       2
       3
       4
       5
       6
       7
       8
       9
      10
      11
      12
      13
      14
      15
      16
      17
      18
      19
      20
      21
      22
      23
      24
      25
      26
      27
      28
      29
      30
      31
      32
      33
      34
      35
      36
      37
      38
      39
      40
      41
      42
      43
      44
      45
      46
      47
      48
      49
      50
      51
      52
      53
      54
      55
      56
      57
      58
      59
      60
      61
      62
      63
      64
      65
      66
      67
      68
      69
      70
      71
      72
      73
      74
      75
      76
      77
      78
      79
      80
      81
      82
      83
      84
      85
      86
      87
      88
      89
      90
      91
      92
      93
      94
      95
      96
      97
      98
      99
     100
     101
     102
     103
     104
     105
     106
     107
     108
     109
     110
     111
     112
     113
     114
     115
     116
     117
     118
     119
     120
     121
     122
     123
     124
     125
     126
     127
     128
     129
     130
     131
     132
     133
     134
     135
     136
     137
     138
     139
     140
     141
     142
     143
     144
     145
     146
     147
     148
     149
     150
     151
      SUBROUTINE CPTCON( N, D, E, ANORM, RCOND, RWORK, INFO )
*
*  -- LAPACK routine (version 3.3.1) --
*  -- LAPACK is a software package provided by Univ. of Tennessee,    --
*  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
*  -- April 2011                                                      --
*
*     .. Scalar Arguments ..
      INTEGER            INFO, N
      REAL               ANORM, RCOND
*     ..
*     .. Array Arguments ..
      REAL               D( * ), RWORK( * )
      COMPLEX            E( * )
*     ..
*
*  Purpose
*  =======
*
*  CPTCON computes the reciprocal of the condition number (in the
*  1-norm) of a complex Hermitian positive definite tridiagonal matrix
*  using the factorization A = L*D*L**H or A = U**H*D*U computed by
*  CPTTRF.
*
*  Norm(inv(A)) is computed by a direct method, and the reciprocal of
*  the condition number is computed as
*                   RCOND = 1 / (ANORM * norm(inv(A))).
*
*  Arguments
*  =========
*
*  N       (input) INTEGER
*          The order of the matrix A.  N >= 0.
*
*  D       (input) REAL array, dimension (N)
*          The n diagonal elements of the diagonal matrix D from the
*          factorization of A, as computed by CPTTRF.
*
*  E       (input) COMPLEX array, dimension (N-1)
*          The (n-1) off-diagonal elements of the unit bidiagonal factor
*          U or L from the factorization of A, as computed by CPTTRF.
*
*  ANORM   (input) REAL
*          The 1-norm of the original matrix A.
*
*  RCOND   (output) REAL
*          The reciprocal of the condition number of the matrix A,
*          computed as RCOND = 1/(ANORM * AINVNM), where AINVNM is the
*          1-norm of inv(A) computed in this routine.
*
*  RWORK   (workspace) REAL array, dimension (N)
*
*  INFO    (output) INTEGER
*          = 0:  successful exit
*          < 0:  if INFO = -i, the i-th argument had an illegal value
*
*  Further Details
*  ===============
*
*  The method used is described in Nicholas J. Higham, "Efficient
*  Algorithms for Computing the Condition Number of a Tridiagonal
*  Matrix", SIAM J. Sci. Stat. Comput., Vol. 7, No. 1, January 1986.
*
*  =====================================================================
*
*     .. Parameters ..
      REAL               ONE, ZERO
      PARAMETER          ( ONE = 1.0E+0, ZERO = 0.0E+0 )
*     ..
*     .. Local Scalars ..
      INTEGER            I, IX
      REAL               AINVNM
*     ..
*     .. External Functions ..
      INTEGER            ISAMAX
      EXTERNAL           ISAMAX
*     ..
*     .. External Subroutines ..
      EXTERNAL           XERBLA
*     ..
*     .. Intrinsic Functions ..
      INTRINSIC          ABS
*     ..
*     .. Executable Statements ..
*
*     Test the input arguments.
*
      INFO = 0
      IF( N.LT.0 ) THEN
         INFO = -1
      ELSE IF( ANORM.LT.ZERO ) THEN
         INFO = -4
      END IF
      IF( INFO.NE.0 ) THEN
         CALL XERBLA( 'CPTCON'-INFO )
         RETURN
      END IF
*
*     Quick return if possible
*
      RCOND = ZERO
      IF( N.EQ.0 ) THEN
         RCOND = ONE
         RETURN
      ELSE IF( ANORM.EQ.ZERO ) THEN
         RETURN
      END IF
*
*     Check that D(1:N) is positive.
*
      DO 10 I = 1, N
         IF( D( I ).LE.ZERO )
     $      RETURN
   10 CONTINUE
*
*     Solve M(A) * x = e, where M(A) = (m(i,j)) is given by
*
*        m(i,j) =  abs(A(i,j)), i = j,
*        m(i,j) = -abs(A(i,j)), i .ne. j,
*
*     and e = [ 1, 1, ..., 1 ]**T.  Note M(A) = M(L)*D*M(L)**H.
*
*     Solve M(L) * x = e.
*
      RWORK( 1 ) = ONE
      DO 20 I = 2, N
         RWORK( I ) = ONE + RWORK( I-1 )*ABS( E( I-1 ) )
   20 CONTINUE
*
*     Solve D * M(L)**H * x = b.
*
      RWORK( N ) = RWORK( N ) / D( N )
      DO 30 I = N - 11-1
         RWORK( I ) = RWORK( I ) / D( I ) + RWORK( I+1 )*ABS( E( I ) )
   30 CONTINUE
*
*     Compute AINVNM = max(x(i)), 1<=i<=n.
*
      IX = ISAMAX( N, RWORK, 1 )
      AINVNM = ABS( RWORK( IX ) )
*
*     Compute the reciprocal condition number.
*
      IF( AINVNM.NE.ZERO )
     $   RCOND = ( ONE / AINVNM ) / ANORM
*
      RETURN
*
*     End of CPTCON
*
      END