SLATEC Routines --- BQR ---


*DECK BQR
      SUBROUTINE BQR (NM, N, MB, A, T, R, IERR, NV, RV)
C***BEGIN PROLOGUE  BQR
C***PURPOSE  Compute some of the eigenvalues of a real symmetric
C            matrix using the QR method with shifts of origin.
C***LIBRARY   SLATEC (EISPACK)
C***CATEGORY  D4A6
C***TYPE      SINGLE PRECISION (BQR-S)
C***KEYWORDS  EIGENVALUES, EISPACK
C***AUTHOR  Smith, B. T., et al.
C***DESCRIPTION
C
C     This subroutine is a translation of the ALGOL procedure BQR,
C     NUM. MATH. 16, 85-92(1970) by Martin, Reinsch, and Wilkinson.
C     HANDBOOK FOR AUTO. COMP., VOL II-LINEAR ALGEBRA, 266-272(1971).
C
C     This subroutine finds the eigenvalue of smallest (usually)
C     magnitude of a REAL SYMMETRIC BAND matrix using the
C     QR algorithm with shifts of origin.  Consecutive calls
C     can be made to find further eigenvalues.
C
C     On INPUT
C
C        NM must be set to the row dimension of the two-dimensional
C          array parameter, A, as declared in the calling program
C          dimension statement.  NM is an INTEGER variable.
C
C        N is the order of the matrix A.  N is an INTEGER variable.
C          N must be less than or equal to NM.
C
C        MB is the (half) band width of the matrix, defined as the
C          number of adjacent diagonals, including the principal
C          diagonal, required to specify the non-zero portion of the
C          lower triangle of the matrix.  MB is an INTEGER variable.
C          MB must be less than or equal to N on first call.
C
C        A contains the lower triangle of the symmetric band input
C          matrix stored as an N by MB array.  Its lowest subdiagonal
C          is stored in the last N+1-MB positions of the first column,
C          its next subdiagonal in the last N+2-MB positions of the
C          second column, further subdiagonals similarly, and finally
C          its principal diagonal in the N positions of the last column.
C          Contents of storages not part of the matrix are arbitrary.
C          On a subsequent call, its output contents from the previous
C          call should be passed.  A is a two-dimensional REAL array,
C          dimensioned A(NM,MB).
C
C        T specifies the shift (of eigenvalues) applied to the diagonal
C          of A in forming the input matrix. What is actually determined
C          is the eigenvalue of A+TI (I is the identity matrix) nearest
C          to T.  On a subsequent call, the output value of T from the
C          previous call should be passed if the next nearest eigenvalue
C          is sought.  T is a REAL variable.
C
C        R should be specified as zero on the first call, and as its
C          output value from the previous call on a subsequent call.
C          It is used to determine when the last row and column of
C          the transformed band matrix can be regarded as negligible.
C          R is a REAL variable.
C
C        NV must be set to the dimension of the array parameter RV
C          as declared in the calling program dimension statement.
C          NV is an INTEGER variable.
C
C     On OUTPUT
C
C        A contains the transformed band matrix.  The matrix A+TI
C          derived from the output parameters is similar to the
C          input A+TI to within rounding errors.  Its last row and
C          column are null (if IERR is zero).
C
C        T contains the computed eigenvalue of A+TI (if IERR is zero),
C          where I is the identity matrix.
C
C        R contains the maximum of its input value and the norm of the
C          last column of the input matrix A.
C
C        IERR is an INTEGER flag set to
C          Zero       for normal return,
C          J          if the J-th eigenvalue has not been
C                     determined after a total of 30 iterations.
C
C        RV is a one-dimensional REAL array of dimension NV which is
C          at least (2*MB**2+4*MB-3), used for temporary storage.  The
C          first (3*MB-2) locations correspond to the ALGOL array B,
C          the next (2*MB-1) locations correspond to the ALGOL array H,
C          and the final (2*MB**2-MB) locations correspond to the MB
C          by (2*MB-1) ALGOL array U.
C
C     NOTE. For a subsequent call, N should be replaced by N-1, but
C     MB should not be altered even when it exceeds the current N.
C
C     Calls PYTHAG(A,B) for SQRT(A**2 + B**2).
C
C     Questions and comments should be directed to B. S. Garbow,
C     Applied Mathematics Division, ARGONNE NATIONAL LABORATORY
C     ------------------------------------------------------------------
C
C***REFERENCES  B. T. Smith, J. M. Boyle, J. J. Dongarra, B. S. Garbow,
C                 Y. Ikebe, V. C. Klema and C. B. Moler, Matrix Eigen-
C                 system Routines - EISPACK Guide, Springer-Verlag,
C                 1976.
C***ROUTINES CALLED  PYTHAG
C***REVISION HISTORY  (YYMMDD)
C   760101  DATE WRITTEN
C   890531  Changed all specific intrinsics to generic.  (WRB)
C   890831  Modified array declarations.  (WRB)
C   890831  REVISION DATE from Version 3.2
C   891214  Prologue converted to Version 4.0 format.  (BAB)
C   920501  Reformatted the REFERENCES section.  (WRB)
C***END PROLOGUE  BQR