NAG Library Manual, Mark 30.1
Interfaces:  FL   CL   CPP   AD 

NAG FL Interface Introduction
Example description
    Program f11drfe

!     F11DRF Example Program Text

!     Mark 30.1 Release. NAG Copyright 2024.

!     .. Use Statements ..
      Use nag_library, Only: f11brf, f11bsf, f11btf, f11drf, f11xnf, nag_wp
!     .. Implicit None Statement ..
      Implicit None
!     .. Parameters ..
      Integer, Parameter               :: nin = 5, nout = 6
!     .. Local Scalars ..
      Real (Kind=nag_wp)               :: anorm, omega, sigmax, stplhs,        &
                                          stprhs, tol
      Integer                          :: i, ifail, ifail1, irevcm, iterm,     &
                                          itn, liwork, lwneed, lwork, m,       &
                                          maxitn, monit, n, nnz
      Character (1)                    :: ckdrf, ckxnf, norm, precon, trans,   &
                                          weight
      Character (8)                    :: method
!     .. Local Arrays ..
      Complex (Kind=nag_wp), Allocatable :: a(:), b(:), rdiag(:), work(:),     &
                                          x(:)
      Real (Kind=nag_wp), Allocatable  :: wgt(:)
      Integer, Allocatable             :: icol(:), irow(:), iwork(:)
!     .. Intrinsic Procedures ..
      Intrinsic                        :: max
!     .. Executable Statements ..
      Write (nout,*) 'F11DRF Example Program Results'
      Write (nout,*)
!     Skip heading in data file
      Read (nin,*)

!     Read algorithmic parameters

      Read (nin,*) n, m
      Read (nin,*) nnz
      lwork = max(121+n*(3+m)+m*(m+5),120+7*n,120+(2*n+m)*(m+2)+2*n,120+10*n)
      liwork = 2*n + 1
      Allocate (a(nnz),b(n),rdiag(n),work(lwork),x(n),wgt(n),icol(nnz),        &
        irow(nnz),iwork(liwork))
      Read (nin,*) method
      Read (nin,*) precon, norm, iterm
      Read (nin,*) tol, maxitn
      Read (nin,*) anorm, sigmax
      Read (nin,*) omega

!     Read the matrix A

      Do i = 1, nnz
        Read (nin,*) a(i), irow(i), icol(i)
      End Do

!     Read rhs vector b and initial approximate solution x

      Read (nin,*) b(1:n)
      Read (nin,*) x(1:n)

!     Call F11BRF to initialize solver

      weight = 'N'
      monit = 0

!     ifail: behaviour on error exit
!             =0 for hard exit, =1 for quiet-soft, =-1 for noisy-soft
      ifail = 0
      Call f11brf(method,precon,norm,weight,iterm,n,m,tol,maxitn,anorm,sigmax, &
        monit,lwneed,work,lwork,ifail)

!     Calculate reciprocal diagonal matrix elements if necessary

      If (precon=='P' .Or. precon=='p') Then

        iwork(1:n) = 0

        Do i = 1, nnz
          If (irow(i)==icol(i)) Then
            iwork(irow(i)) = iwork(irow(i)) + 1
            If (a(i)/=(0.0E0_nag_wp,0.0E0_nag_wp)) Then
              rdiag(irow(i)) = (1.0E0_nag_wp,0.0E0_nag_wp)/a(i)
            Else
              Write (nout,*) 'Matrix has a zero diagonal element'
              Go To 100
            End If
          End If
        End Do

        Do i = 1, n
          If (iwork(i)==0) Then
            Write (nout,*) 'Matrix has a missing diagonal element'
            Go To 100
          End If
          If (iwork(i)>=2) Then
            Write (nout,*) 'Matrix has a multiple diagonal element'
            Go To 100
          End If
        End Do

      End If

!     Call F11BSF to solve the linear system

      irevcm = 0
      ckxnf = 'C'
      ckdrf = 'C'

      ifail = 1
loop: Do
        Call f11bsf(irevcm,x,b,wgt,work,lwork,ifail)

        If (irevcm/=4) Then
          ifail1 = 1
          Select Case (irevcm)
          Case (1)
!           Compute matrix-vector product
            trans = 'N'

            Call f11xnf(trans,n,nnz,a,irow,icol,ckxnf,x,b,ifail1)

            ckxnf = 'N'
          Case (-1)
!           Compute conjugate transposed matrix-vector product
            trans = 'T'

            Call f11xnf(trans,n,nnz,a,irow,icol,ckxnf,x,b,ifail1)

            ckxnf = 'N'
          Case (2)
!           SSOR preconditioning
            trans = 'N'

            Call f11drf(trans,n,nnz,a,irow,icol,rdiag,omega,ckdrf,x,b,iwork,   &
              ifail1)

            ckdrf = 'N'
          End Select
          If (ifail1/=0) Then
            irevcm = 6
          End If
        Else If (ifail==0) Then
!         Termination

          ifail = 0
          Call f11btf(itn,stplhs,stprhs,anorm,sigmax,work,lwork,ifail)

          Write (nout,99996) itn
          Write (nout,99997) 'Matrix norm =', anorm
          Write (nout,99997) 'Final residual norm =', stplhs
          Write (nout,*)

!         Output x
          Write (nout,*) '                   X'
          Write (nout,99998) x(1:n)

          Exit loop
        Else
          Write (nout,99999) ifail
          Exit loop
        End If
      End Do loop
100   Continue

99999 Format (1X,/,1X,' ** F11BSF returned with IFAIL = ',I5)
99998 Format (1X,'(',1P,E16.4,',',1P,E16.4,')')
99997 Format (1X,A,1P,E16.3)
99996 Format (1X,'Converged in',I10,' iterations')
    End Program f11drfe