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 |
SUBROUTINE CTBT06( RCOND, RCONDC, UPLO, DIAG, N, KD, AB, LDAB,
$ RWORK, RAT ) * * -- LAPACK test routine (version 3.1) -- * Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd.. * November 2006 * * .. Scalar Arguments .. CHARACTER DIAG, UPLO INTEGER KD, LDAB, N REAL RAT, RCOND, RCONDC * .. * .. Array Arguments .. REAL RWORK( * ) COMPLEX AB( LDAB, * ) * .. * * Purpose * ======= * * CTBT06 computes a test ratio comparing RCOND (the reciprocal * condition number of a triangular matrix A) and RCONDC, the estimate * computed by CTBCON. Information about the triangular matrix A is * used if one estimate is zero and the other is non-zero to decide if * underflow in the estimate is justified. * * Arguments * ========= * * RCOND (input) REAL * The estimate of the reciprocal condition number obtained by * forming the explicit inverse of the matrix A and computing * RCOND = 1/( norm(A) * norm(inv(A)) ). * * RCONDC (input) REAL * The estimate of the reciprocal condition number computed by * CTBCON. * * UPLO (input) CHARACTER * Specifies whether the matrix A is upper or lower triangular. * = 'U': Upper triangular * = 'L': Lower triangular * * DIAG (input) CHARACTER * Specifies whether or not the matrix A is unit triangular. * = 'N': Non-unit triangular * = 'U': Unit triangular * * N (input) INTEGER * The order of the matrix A. N >= 0. * * KD (input) INTEGER * The number of superdiagonals or subdiagonals of the * triangular band matrix A. KD >= 0. * * AB (input) COMPLEX array, dimension (LDAB,N) * The upper or lower triangular band matrix A, stored in the * first kd+1 rows of the array. The j-th column of A is stored * in the j-th column of the array AB as follows: * if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j; * if UPLO = 'L', AB(1+i-j,j) = A(i,j) for j<=i<=min(n,j+kd). * * LDAB (input) INTEGER * The leading dimension of the array AB. LDAB >= KD+1. * * RWORK (workspace) REAL array, dimension (N) * * RAT (output) REAL * The test ratio. If both RCOND and RCONDC are nonzero, * RAT = MAX( RCOND, RCONDC )/MIN( RCOND, RCONDC ) - 1. * If RAT = 0, the two estimates are exactly the same. * * ===================================================================== * * .. Parameters .. REAL ZERO, ONE PARAMETER ( ZERO = 0.0E+0, ONE = 1.0E+0 ) * .. * .. Local Scalars .. REAL ANORM, BIGNUM, EPS, RMAX, RMIN * .. * .. External Functions .. REAL CLANTB, SLAMCH EXTERNAL CLANTB, SLAMCH * .. * .. Intrinsic Functions .. INTRINSIC MAX, MIN * .. * .. Executable Statements .. * EPS = SLAMCH( 'Epsilon' ) RMAX = MAX( RCOND, RCONDC ) RMIN = MIN( RCOND, RCONDC ) * * Do the easy cases first. * IF( RMIN.LT.ZERO ) THEN * * Invalid value for RCOND or RCONDC, return 1/EPS. * RAT = ONE / EPS * ELSE IF( RMIN.GT.ZERO ) THEN * * Both estimates are positive, return RMAX/RMIN - 1. * RAT = RMAX / RMIN - ONE * ELSE IF( RMAX.EQ.ZERO ) THEN * * Both estimates zero. * RAT = ZERO * ELSE * * One estimate is zero, the other is non-zero. If the matrix is * ill-conditioned, return the nonzero estimate multiplied by * 1/EPS; if the matrix is badly scaled, return the nonzero * estimate multiplied by BIGNUM/TMAX, where TMAX is the maximum * element in absolute value in A. * BIGNUM = ONE / SLAMCH( 'Safe minimum' ) ANORM = CLANTB( 'M', UPLO, DIAG, N, KD, AB, LDAB, RWORK ) * RAT = RMAX*( MIN( BIGNUM / MAX( ONE, ANORM ), ONE / EPS ) ) END IF * RETURN * * End of CTBT06 * END |