LAPACK  3.10.0
LAPACK: Linear Algebra PACKage

◆ cpbsvx()

subroutine cpbsvx ( character  FACT,
character  UPLO,
integer  N,
integer  KD,
integer  NRHS,
complex, dimension( ldab, * )  AB,
integer  LDAB,
complex, dimension( ldafb, * )  AFB,
integer  LDAFB,
character  EQUED,
real, dimension( * )  S,
complex, dimension( ldb, * )  B,
integer  LDB,
complex, dimension( ldx, * )  X,
integer  LDX,
real  RCOND,
real, dimension( * )  FERR,
real, dimension( * )  BERR,
complex, dimension( * )  WORK,
real, dimension( * )  RWORK,
integer  INFO 
)

CPBSVX computes the solution to system of linear equations A * X = B for OTHER matrices

Download CPBSVX + dependencies [TGZ] [ZIP] [TXT]

Purpose:
 CPBSVX uses the Cholesky factorization A = U**H*U or A = L*L**H to
 compute the solution to a complex system of linear equations
    A * X = B,
 where A is an N-by-N Hermitian positive definite band matrix and X
 and B are N-by-NRHS matrices.

 Error bounds on the solution and a condition estimate are also
 provided.
Description:
 The following steps are performed:

 1. If FACT = 'E', real scaling factors are computed to equilibrate
    the system:
       diag(S) * A * diag(S) * inv(diag(S)) * X = diag(S) * B
    Whether or not the system will be equilibrated depends on the
    scaling of the matrix A, but if equilibration is used, A is
    overwritten by diag(S)*A*diag(S) and B by diag(S)*B.

 2. If FACT = 'N' or 'E', the Cholesky decomposition is used to
    factor the matrix A (after equilibration if FACT = 'E') as
       A = U**H * U,  if UPLO = 'U', or
       A = L * L**H,  if UPLO = 'L',
    where U is an upper triangular band matrix, and L is a lower
    triangular band matrix.

 3. If the leading i-by-i principal minor is not positive definite,
    then the routine returns with INFO = i. Otherwise, the factored
    form of A is used to estimate the condition number of the matrix
    A.  If the reciprocal of the condition number is less than machine
    precision, INFO = N+1 is returned as a warning, but the routine
    still goes on to solve for X and compute error bounds as
    described below.

 4. The system of equations is solved for X using the factored form
    of A.

 5. Iterative refinement is applied to improve the computed solution
    matrix and calculate error bounds and backward error estimates
    for it.

 6. If equilibration was used, the matrix X is premultiplied by
    diag(S) so that it solves the original system before
    equilibration.
Parameters
[in]FACT
          FACT is CHARACTER*1
          Specifies whether or not the factored form of the matrix A is
          supplied on entry, and if not, whether the matrix A should be
          equilibrated before it is factored.
          = 'F':  On entry, AFB contains the factored form of A.
                  If EQUED = 'Y', the matrix A has been equilibrated
                  with scaling factors given by S.  AB and AFB will not
                  be modified.
          = 'N':  The matrix A will be copied to AFB and factored.
          = 'E':  The matrix A will be equilibrated if necessary, then
                  copied to AFB and factored.
[in]UPLO
          UPLO is CHARACTER*1
          = 'U':  Upper triangle of A is stored;
          = 'L':  Lower triangle of A is stored.
[in]N
          N is INTEGER
          The number of linear equations, i.e., the order of the
          matrix A.  N >= 0.
[in]KD
          KD is INTEGER
          The number of superdiagonals of the matrix A if UPLO = 'U',
          or the number of subdiagonals if UPLO = 'L'.  KD >= 0.
[in]NRHS
          NRHS is INTEGER
          The number of right-hand sides, i.e., the number of columns
          of the matrices B and X.  NRHS >= 0.
[in,out]AB
          AB is COMPLEX array, dimension (LDAB,N)
          On entry, the upper or lower triangle of the Hermitian band
          matrix A, stored in the first KD+1 rows of the array, except
          if FACT = 'F' and EQUED = 'Y', then A must contain the
          equilibrated matrix diag(S)*A*diag(S).  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).
          See below for further details.

          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
          diag(S)*A*diag(S).
[in]LDAB
          LDAB is INTEGER
          The leading dimension of the array A.  LDAB >= KD+1.
[in,out]AFB
          AFB is COMPLEX array, dimension (LDAFB,N)
          If FACT = 'F', then AFB is an input argument and on entry
          contains the triangular factor U or L from the Cholesky
          factorization A = U**H*U or A = L*L**H of the band matrix
          A, in the same storage format as A (see AB).  If EQUED = 'Y',
          then AFB is the factored form of the equilibrated matrix A.

          If FACT = 'N', then AFB is an output argument and on exit
          returns the triangular factor U or L from the Cholesky
          factorization A = U**H*U or A = L*L**H.

          If FACT = 'E', then AFB is an output argument and on exit
          returns the triangular factor U or L from the Cholesky
          factorization A = U**H*U or A = L*L**H of the equilibrated
          matrix A (see the description of A for the form of the
          equilibrated matrix).
[in]LDAFB
          LDAFB is INTEGER
          The leading dimension of the array AFB.  LDAFB >= KD+1.
[in,out]EQUED
          EQUED is CHARACTER*1
          Specifies the form of equilibration that was done.
          = 'N':  No equilibration (always true if FACT = 'N').
          = 'Y':  Equilibration was done, i.e., A has been replaced by
                  diag(S) * A * diag(S).
          EQUED is an input argument if FACT = 'F'; otherwise, it is an
          output argument.
[in,out]S
          S is REAL array, dimension (N)
          The scale factors for A; not accessed if EQUED = 'N'.  S is
          an input argument if FACT = 'F'; otherwise, S is an output
          argument.  If FACT = 'F' and EQUED = 'Y', each element of S
          must be positive.
[in,out]B
          B is COMPLEX array, dimension (LDB,NRHS)
          On entry, the N-by-NRHS right hand side matrix B.
          On exit, if EQUED = 'N', B is not modified; if EQUED = 'Y',
          B is overwritten by diag(S) * B.
[in]LDB
          LDB is INTEGER
          The leading dimension of the array B.  LDB >= max(1,N).
[out]X
          X is COMPLEX array, dimension (LDX,NRHS)
          If INFO = 0 or INFO = N+1, the N-by-NRHS solution matrix X to
          the original system of equations.  Note that if EQUED = 'Y',
          A and B are modified on exit, and the solution to the
          equilibrated system is inv(diag(S))*X.
[in]LDX
          LDX is INTEGER
          The leading dimension of the array X.  LDX >= max(1,N).
[out]RCOND
          RCOND is REAL
          The estimate of the reciprocal condition number of the matrix
          A after equilibration (if done).  If RCOND is less than the
          machine precision (in particular, if RCOND = 0), the matrix
          is singular to working precision.  This condition is
          indicated by a return code of INFO > 0.
[out]FERR
          FERR is REAL array, dimension (NRHS)
          The estimated forward error bound for each solution vector
          X(j) (the j-th column of the solution matrix X).
          If XTRUE is the true solution corresponding to X(j), FERR(j)
          is an estimated upper bound for the magnitude of the largest
          element in (X(j) - XTRUE) divided by the magnitude of the
          largest element in X(j).  The estimate is as reliable as
          the estimate for RCOND, and is almost always a slight
          overestimate of the true error.
[out]BERR
          BERR is REAL array, dimension (NRHS)
          The componentwise relative backward error of each solution
          vector X(j) (i.e., the smallest relative change in
          any element of A or B that makes X(j) an exact solution).
[out]WORK
          WORK is COMPLEX array, dimension (2*N)
[out]RWORK
          RWORK is REAL array, dimension (N)
[out]INFO
          INFO is INTEGER
          = 0: successful exit
          < 0: if INFO = -i, the i-th argument had an illegal value
          > 0: if INFO = i, and i is
                <= N:  the leading minor of order i of A is
                       not positive definite, so the factorization
                       could not be completed, and the solution has not
                       been computed. RCOND = 0 is returned.
                = N+1: U is nonsingular, but RCOND is less than machine
                       precision, meaning that the matrix is singular
                       to working precision.  Nevertheless, the
                       solution and error bounds are computed because
                       there are a number of situations where the
                       computed solution can be more accurate than the
                       value of RCOND would suggest.
Author
Univ. of Tennessee
Univ. of California Berkeley
Univ. of Colorado Denver
NAG Ltd.
Further Details:
  The band storage scheme is illustrated by the following example, when
  N = 6, KD = 2, and UPLO = 'U':

  Two-dimensional storage of the Hermitian matrix A:

     a11  a12  a13
          a22  a23  a24
               a33  a34  a35
                    a44  a45  a46
                         a55  a56
     (aij=conjg(aji))         a66

  Band storage of the upper triangle of A:

      *    *   a13  a24  a35  a46
      *   a12  a23  a34  a45  a56
     a11  a22  a33  a44  a55  a66

  Similarly, if UPLO = 'L' the format of A is as follows:

     a11  a22  a33  a44  a55  a66
     a21  a32  a43  a54  a65   *
     a31  a42  a53  a64   *    *

  Array elements marked * are not used by the routine.

Definition at line 339 of file cpbsvx.f.

342 *
343 * -- LAPACK driver routine --
344 * -- LAPACK is a software package provided by Univ. of Tennessee, --
345 * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
346 *
347 * .. Scalar Arguments ..
348  CHARACTER EQUED, FACT, UPLO
349  INTEGER INFO, KD, LDAB, LDAFB, LDB, LDX, N, NRHS
350  REAL RCOND
351 * ..
352 * .. Array Arguments ..
353  REAL BERR( * ), FERR( * ), RWORK( * ), S( * )
354  COMPLEX AB( LDAB, * ), AFB( LDAFB, * ), B( LDB, * ),
355  $ WORK( * ), X( LDX, * )
356 * ..
357 *
358 * =====================================================================
359 *
360 * .. Parameters ..
361  REAL ZERO, ONE
362  parameter( zero = 0.0e+0, one = 1.0e+0 )
363 * ..
364 * .. Local Scalars ..
365  LOGICAL EQUIL, NOFACT, RCEQU, UPPER
366  INTEGER I, INFEQU, J, J1, J2
367  REAL AMAX, ANORM, BIGNUM, SCOND, SMAX, SMIN, SMLNUM
368 * ..
369 * .. External Functions ..
370  LOGICAL LSAME
371  REAL CLANHB, SLAMCH
372  EXTERNAL lsame, clanhb, slamch
373 * ..
374 * .. External Subroutines ..
375  EXTERNAL ccopy, clacpy, claqhb, cpbcon, cpbequ, cpbrfs,
376  $ cpbtrf, cpbtrs, xerbla
377 * ..
378 * .. Intrinsic Functions ..
379  INTRINSIC max, min
380 * ..
381 * .. Executable Statements ..
382 *
383  info = 0
384  nofact = lsame( fact, 'N' )
385  equil = lsame( fact, 'E' )
386  upper = lsame( uplo, 'U' )
387  IF( nofact .OR. equil ) THEN
388  equed = 'N'
389  rcequ = .false.
390  ELSE
391  rcequ = lsame( equed, 'Y' )
392  smlnum = slamch( 'Safe minimum' )
393  bignum = one / smlnum
394  END IF
395 *
396 * Test the input parameters.
397 *
398  IF( .NOT.nofact .AND. .NOT.equil .AND. .NOT.lsame( fact, 'F' ) )
399  $ THEN
400  info = -1
401  ELSE IF( .NOT.upper .AND. .NOT.lsame( uplo, 'L' ) ) THEN
402  info = -2
403  ELSE IF( n.LT.0 ) THEN
404  info = -3
405  ELSE IF( kd.LT.0 ) THEN
406  info = -4
407  ELSE IF( nrhs.LT.0 ) THEN
408  info = -5
409  ELSE IF( ldab.LT.kd+1 ) THEN
410  info = -7
411  ELSE IF( ldafb.LT.kd+1 ) THEN
412  info = -9
413  ELSE IF( lsame( fact, 'F' ) .AND. .NOT.
414  $ ( rcequ .OR. lsame( equed, 'N' ) ) ) THEN
415  info = -10
416  ELSE
417  IF( rcequ ) THEN
418  smin = bignum
419  smax = zero
420  DO 10 j = 1, n
421  smin = min( smin, s( j ) )
422  smax = max( smax, s( j ) )
423  10 CONTINUE
424  IF( smin.LE.zero ) THEN
425  info = -11
426  ELSE IF( n.GT.0 ) THEN
427  scond = max( smin, smlnum ) / min( smax, bignum )
428  ELSE
429  scond = one
430  END IF
431  END IF
432  IF( info.EQ.0 ) THEN
433  IF( ldb.LT.max( 1, n ) ) THEN
434  info = -13
435  ELSE IF( ldx.LT.max( 1, n ) ) THEN
436  info = -15
437  END IF
438  END IF
439  END IF
440 *
441  IF( info.NE.0 ) THEN
442  CALL xerbla( 'CPBSVX', -info )
443  RETURN
444  END IF
445 *
446  IF( equil ) THEN
447 *
448 * Compute row and column scalings to equilibrate the matrix A.
449 *
450  CALL cpbequ( uplo, n, kd, ab, ldab, s, scond, amax, infequ )
451  IF( infequ.EQ.0 ) THEN
452 *
453 * Equilibrate the matrix.
454 *
455  CALL claqhb( uplo, n, kd, ab, ldab, s, scond, amax, equed )
456  rcequ = lsame( equed, 'Y' )
457  END IF
458  END IF
459 *
460 * Scale the right-hand side.
461 *
462  IF( rcequ ) THEN
463  DO 30 j = 1, nrhs
464  DO 20 i = 1, n
465  b( i, j ) = s( i )*b( i, j )
466  20 CONTINUE
467  30 CONTINUE
468  END IF
469 *
470  IF( nofact .OR. equil ) THEN
471 *
472 * Compute the Cholesky factorization A = U**H *U or A = L*L**H.
473 *
474  IF( upper ) THEN
475  DO 40 j = 1, n
476  j1 = max( j-kd, 1 )
477  CALL ccopy( j-j1+1, ab( kd+1-j+j1, j ), 1,
478  $ afb( kd+1-j+j1, j ), 1 )
479  40 CONTINUE
480  ELSE
481  DO 50 j = 1, n
482  j2 = min( j+kd, n )
483  CALL ccopy( j2-j+1, ab( 1, j ), 1, afb( 1, j ), 1 )
484  50 CONTINUE
485  END IF
486 *
487  CALL cpbtrf( uplo, n, kd, afb, ldafb, info )
488 *
489 * Return if INFO is non-zero.
490 *
491  IF( info.GT.0 )THEN
492  rcond = zero
493  RETURN
494  END IF
495  END IF
496 *
497 * Compute the norm of the matrix A.
498 *
499  anorm = clanhb( '1', uplo, n, kd, ab, ldab, rwork )
500 *
501 * Compute the reciprocal of the condition number of A.
502 *
503  CALL cpbcon( uplo, n, kd, afb, ldafb, anorm, rcond, work, rwork,
504  $ info )
505 *
506 * Compute the solution matrix X.
507 *
508  CALL clacpy( 'Full', n, nrhs, b, ldb, x, ldx )
509  CALL cpbtrs( uplo, n, kd, nrhs, afb, ldafb, x, ldx, info )
510 *
511 * Use iterative refinement to improve the computed solution and
512 * compute error bounds and backward error estimates for it.
513 *
514  CALL cpbrfs( uplo, n, kd, nrhs, ab, ldab, afb, ldafb, b, ldb, x,
515  $ ldx, ferr, berr, work, rwork, info )
516 *
517 * Transform the solution matrix X to a solution of the original
518 * system.
519 *
520  IF( rcequ ) THEN
521  DO 70 j = 1, nrhs
522  DO 60 i = 1, n
523  x( i, j ) = s( i )*x( i, j )
524  60 CONTINUE
525  70 CONTINUE
526  DO 80 j = 1, nrhs
527  ferr( j ) = ferr( j ) / scond
528  80 CONTINUE
529  END IF
530 *
531 * Set INFO = N+1 if the matrix is singular to working precision.
532 *
533  IF( rcond.LT.slamch( 'Epsilon' ) )
534  $ info = n + 1
535 *
536  RETURN
537 *
538 * End of CPBSVX
539 *
subroutine xerbla(SRNAME, INFO)
XERBLA
Definition: xerbla.f:60
logical function lsame(CA, CB)
LSAME
Definition: lsame.f:53
subroutine ccopy(N, CX, INCX, CY, INCY)
CCOPY
Definition: ccopy.f:81
subroutine claqhb(UPLO, N, KD, AB, LDAB, S, SCOND, AMAX, EQUED)
CLAQHB scales a Hermitian band matrix, using scaling factors computed by cpbequ.
Definition: claqhb.f:141
real function clanhb(NORM, UPLO, N, K, AB, LDAB, WORK)
CLANHB returns the value of the 1-norm, or the Frobenius norm, or the infinity norm,...
Definition: clanhb.f:132
subroutine clacpy(UPLO, M, N, A, LDA, B, LDB)
CLACPY copies all or part of one two-dimensional array to another.
Definition: clacpy.f:103
subroutine cpbequ(UPLO, N, KD, AB, LDAB, S, SCOND, AMAX, INFO)
CPBEQU
Definition: cpbequ.f:130
subroutine cpbrfs(UPLO, N, KD, NRHS, AB, LDAB, AFB, LDAFB, B, LDB, X, LDX, FERR, BERR, WORK, RWORK, INFO)
CPBRFS
Definition: cpbrfs.f:189
subroutine cpbcon(UPLO, N, KD, AB, LDAB, ANORM, RCOND, WORK, RWORK, INFO)
CPBCON
Definition: cpbcon.f:133
subroutine cpbtrs(UPLO, N, KD, NRHS, AB, LDAB, B, LDB, INFO)
CPBTRS
Definition: cpbtrs.f:121
subroutine cpbtrf(UPLO, N, KD, AB, LDAB, INFO)
CPBTRF
Definition: cpbtrf.f:142
real function slamch(CMACH)
SLAMCH
Definition: slamch.f:68
Here is the call graph for this function:
Here is the caller graph for this function: