LAPACK 3.3.0

stpt02.f

Go to the documentation of this file.
00001       SUBROUTINE STPT02( UPLO, TRANS, DIAG, N, NRHS, AP, X, LDX, B, LDB,
00002      $                   WORK, RESID )
00003 *
00004 *  -- LAPACK test routine (version 3.1) --
00005 *     Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd..
00006 *     November 2006
00007 *
00008 *     .. Scalar Arguments ..
00009       CHARACTER          DIAG, TRANS, UPLO
00010       INTEGER            LDB, LDX, N, NRHS
00011       REAL               RESID
00012 *     ..
00013 *     .. Array Arguments ..
00014       REAL               AP( * ), B( LDB, * ), WORK( * ), X( LDX, * )
00015 *     ..
00016 *
00017 *  Purpose
00018 *  =======
00019 *
00020 *  STPT02 computes the residual for the computed solution to a
00021 *  triangular system of linear equations  A*x = b  or  A'*x = b  when
00022 *  the triangular matrix A is stored in packed format.  Here A' is the
00023 *  transpose of A and x and b are N by NRHS matrices.  The test ratio is
00024 *  the maximum over the number of right hand sides of
00025 *     norm(b - op(A)*x) / ( norm(op(A)) * norm(x) * EPS ),
00026 *  where op(A) denotes A or A' and EPS is the machine epsilon.
00027 *
00028 *  Arguments
00029 *  =========
00030 *
00031 *  UPLO    (input) CHARACTER*1
00032 *          Specifies whether the matrix A is upper or lower triangular.
00033 *          = 'U':  Upper triangular
00034 *          = 'L':  Lower triangular
00035 *
00036 *  TRANS   (input) CHARACTER*1
00037 *          Specifies the operation applied to A.
00038 *          = 'N':  A *x = b  (No transpose)
00039 *          = 'T':  A'*x = b  (Transpose)
00040 *          = 'C':  A'*x = b  (Conjugate transpose = Transpose)
00041 *
00042 *  DIAG    (input) CHARACTER*1
00043 *          Specifies whether or not the matrix A is unit triangular.
00044 *          = 'N':  Non-unit triangular
00045 *          = 'U':  Unit triangular
00046 *
00047 *  N       (input) INTEGER
00048 *          The order of the matrix A.  N >= 0.
00049 *
00050 *  NRHS    (input) INTEGER
00051 *          The number of right hand sides, i.e., the number of columns
00052 *          of the matrices X and B.  NRHS >= 0.
00053 *
00054 *  AP      (input) REAL array, dimension (N*(N+1)/2)
00055 *          The upper or lower triangular matrix A, packed columnwise in
00056 *          a linear array.  The j-th column of A is stored in the array
00057 *          AP as follows:
00058 *          if UPLO = 'U', AP((j-1)*j/2 + i) = A(i,j) for 1<=i<=j;
00059 *          if UPLO = 'L',
00060 *             AP((j-1)*(n-j) + j*(j+1)/2 + i-j) = A(i,j) for j<=i<=n.
00061 *
00062 *  X       (input) REAL array, dimension (LDX,NRHS)
00063 *          The computed solution vectors for the system of linear
00064 *          equations.
00065 *
00066 *  LDX     (input) INTEGER
00067 *          The leading dimension of the array X.  LDX >= max(1,N).
00068 *
00069 *  B       (input) REAL array, dimension (LDB,NRHS)
00070 *          The right hand side vectors for the system of linear
00071 *          equations.
00072 *
00073 *  LDB     (input) INTEGER
00074 *          The leading dimension of the array B.  LDB >= max(1,N).
00075 *
00076 *  WORK    (workspace) REAL array, dimension (N)
00077 *
00078 *  RESID   (output) REAL
00079 *          The maximum over the number of right hand sides of
00080 *          norm(op(A)*x - b) / ( norm(op(A)) * norm(x) * EPS ).
00081 *
00082 *  =====================================================================
00083 *
00084 *     .. Parameters ..
00085       REAL               ZERO, ONE
00086       PARAMETER          ( ZERO = 0.0E+0, ONE = 1.0E+0 )
00087 *     ..
00088 *     .. Local Scalars ..
00089       INTEGER            J
00090       REAL               ANORM, BNORM, EPS, XNORM
00091 *     ..
00092 *     .. External Functions ..
00093       LOGICAL            LSAME
00094       REAL               SASUM, SLAMCH, SLANTP
00095       EXTERNAL           LSAME, SASUM, SLAMCH, SLANTP
00096 *     ..
00097 *     .. External Subroutines ..
00098       EXTERNAL           SAXPY, SCOPY, STPMV
00099 *     ..
00100 *     .. Intrinsic Functions ..
00101       INTRINSIC          MAX
00102 *     ..
00103 *     .. Executable Statements ..
00104 *
00105 *     Quick exit if N = 0 or NRHS = 0
00106 *
00107       IF( N.LE.0 .OR. NRHS.LE.0 ) THEN
00108          RESID = ZERO
00109          RETURN
00110       END IF
00111 *
00112 *     Compute the 1-norm of A or A'.
00113 *
00114       IF( LSAME( TRANS, 'N' ) ) THEN
00115          ANORM = SLANTP( '1', UPLO, DIAG, N, AP, WORK )
00116       ELSE
00117          ANORM = SLANTP( 'I', UPLO, DIAG, N, AP, WORK )
00118       END IF
00119 *
00120 *     Exit with RESID = 1/EPS if ANORM = 0.
00121 *
00122       EPS = SLAMCH( 'Epsilon' )
00123       IF( ANORM.LE.ZERO ) THEN
00124          RESID = ONE / EPS
00125          RETURN
00126       END IF
00127 *
00128 *     Compute the maximum over the number of right hand sides of
00129 *        norm(op(A)*x - b) / ( norm(op(A)) * norm(x) * EPS ).
00130 *
00131       RESID = ZERO
00132       DO 10 J = 1, NRHS
00133          CALL SCOPY( N, X( 1, J ), 1, WORK, 1 )
00134          CALL STPMV( UPLO, TRANS, DIAG, N, AP, WORK, 1 )
00135          CALL SAXPY( N, -ONE, B( 1, J ), 1, WORK, 1 )
00136          BNORM = SASUM( N, WORK, 1 )
00137          XNORM = SASUM( N, X( 1, J ), 1 )
00138          IF( XNORM.LE.ZERO ) THEN
00139             RESID = ONE / EPS
00140          ELSE
00141             RESID = MAX( RESID, ( ( BNORM / ANORM ) / XNORM ) / EPS )
00142          END IF
00143    10 CONTINUE
00144 *
00145       RETURN
00146 *
00147 *     End of STPT02
00148 *
00149       END
 All Files Functions