Home Browse Files
       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
      SUBROUTINE SLAE2( A, B, C, RT1, RT2 )
*
*  -- LAPACK auxiliary routine (version 3.2) --
*  -- LAPACK is a software package provided by Univ. of Tennessee,    --
*  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
*     November 2006
*
*     .. Scalar Arguments ..
      REAL               A, B, C, RT1, RT2
*     ..
*
*  Purpose
*  =======
*
*  SLAE2  computes the eigenvalues of a 2-by-2 symmetric matrix
*     [  A   B  ]
*     [  B   C  ].
*  On return, RT1 is the eigenvalue of larger absolute value, and RT2
*  is the eigenvalue of smaller absolute value.
*
*  Arguments
*  =========
*
*  A       (input) REAL
*          The (1,1) element of the 2-by-2 matrix.
*
*  B       (input) REAL
*          The (1,2) and (2,1) elements of the 2-by-2 matrix.
*
*  C       (input) REAL
*          The (2,2) element of the 2-by-2 matrix.
*
*  RT1     (output) REAL
*          The eigenvalue of larger absolute value.
*
*  RT2     (output) REAL
*          The eigenvalue of smaller absolute value.
*
*  Further Details
*  ===============
*
*  RT1 is accurate to a few ulps barring over/underflow.
*
*  RT2 may be inaccurate if there is massive cancellation in the
*  determinant A*C-B*B; higher precision or correctly rounded or
*  correctly truncated arithmetic would be needed to compute RT2
*  accurately in all cases.
*
*  Overflow is possible only if RT1 is within a factor of 5 of overflow.
*  Underflow is harmless if the input data is 0 or exceeds
*     underflow_threshold / macheps.
*
* =====================================================================
*
*     .. Parameters ..
      REAL               ONE
      PARAMETER          ( ONE = 1.0E0 )
      REAL               TWO
      PARAMETER          ( TWO = 2.0E0 )
      REAL               ZERO
      PARAMETER          ( ZERO = 0.0E0 )
      REAL               HALF
      PARAMETER          ( HALF = 0.5E0 )
*     ..
*     .. Local Scalars ..
      REAL               AB, ACMN, ACMX, ADF, DF, RT, SM, TB
*     ..
*     .. Intrinsic Functions ..
      INTRINSIC          ABS, SQRT
*     ..
*     .. Executable Statements ..
*
*     Compute the eigenvalues
*
      SM = A + C
      DF = A - C
      ADF = ABS( DF )
      TB = B + B
      AB = ABS( TB )
      IF( ABS( A ).GT.ABS( C ) ) THEN
         ACMX = A
         ACMN = C
      ELSE
         ACMX = C
         ACMN = A
      END IF
      IF( ADF.GT.AB ) THEN
         RT = ADF*SQRT( ONE+( AB / ADF )**2 )
      ELSE IF( ADF.LT.AB ) THEN
         RT = AB*SQRT( ONE+( ADF / AB )**2 )
      ELSE
*
*        Includes case AB=ADF=0
*
         RT = AB*SQRT( TWO )
      END IF
      IF( SM.LT.ZERO ) THEN
         RT1 = HALF*( SM-RT )
*
*        Order of execution important.
*        To get fully accurate smaller eigenvalue,
*        next line needs to be executed in higher precision.
*
         RT2 = ( ACMX / RT1 )*ACMN - ( B / RT1 )*B
      ELSE IF( SM.GT.ZERO ) THEN
         RT1 = HALF*( SM+RT )
*
*        Order of execution important.
*        To get fully accurate smaller eigenvalue,
*        next line needs to be executed in higher precision.
*
         RT2 = ( ACMX / RT1 )*ACMN - ( B / RT1 )*B
      ELSE
*
*        Includes case RT1 = RT2 = 0
*
         RT1 = HALF*RT
         RT2 = -HALF*RT
      END IF
      RETURN
*
*     End of SLAE2
*
      END





































































































Home