z_bicg z_bicg_r z_cgs z_cgs_r z_cocg z_cocg_r z_cocr z_cocr_r z_diom z_diom_r z_dqgmres z_dqgmres_r z_fgmres z_fgmres_r z_fom z_fom_r z_gcr z_gcr_r z_gpbicg z_gpbicg_r z_orthomin z_orthomin_r z_qmr z_qmr_r z_sor z_sor_r z_tfqmr z_tfqmr_r

z_qmr() void z_qmr ( int  n, void(*)(int, const doublecomplex[], doublecomplex[])  matvec, void(*)(int, const doublecomplex[], doublecomplex[])  matvectrans, void(*)(int, const doublecomplex[], doublecomplex[])  psolve, void(*)(int, const doublecomplex[], doublecomplex[])  psolvetrans, void(*)(int, const doublecomplex[], double, int, int *)  chkconv, const doublecomplex  b[], doublecomplex  x[], int  maxiter, int *  iter, double *  res, int  lwork, doublecomplex  work[], int *  info  ) Solution of linear system Ax = b using quasi minimum residual (QMR) method (Complex matrices) Purpose This routine solves the linear system Ax = b using the quasi minimum residual (QMR) iterative method with preconditioning. Parameters [in] n Dimension of the matrix. (n >= 0) (if n = 0, returns without computation) [in] matvec User supplied subroutine which calculates the matrix and vector product as follows. void matvec(int n, const doublecomplex x[], doublecomplex y[]) { Compute A*x, and return result in y[]. } [in] matvectrans User supplied subroutine which calculates the conjugate transposed matrix and vector product as follows. void matvectrans(int n, const doublecomplex x[], doublecomplex y[]) { Compute A**H*x, and return result in y[]. } [in] psolve User supplied subroutine which perform the preconditioner solve routine for the linear system M*x = b as follows, where M is a preconditioner metrix. void psolve(int n, const doublecomplex b[], doublecomplex x[]) { Solve M*x = b, and return solution in x[]. } [in] psolvetrans User supplied subroutine which perform the preconditioner solve routine for the linear system M**H*x = b as follows, where M is a preconditioner metrix. void psolvetrans(int n, const doublecomplex b[], doublecomplex x[]) { Solve M**H*x = b, and return solution in x[]. } [in] chkconv User supplied subroutine which is called on every iteration for the convergence test as follows, where x[] is the current approximate solution, res is the current residual norm norm(b - A*x), and iter is the current number of iterations. This routine can also be used to output the intermediate results. void chkconv(int n, const doublecomplex x[], double res, int iter, int *ichk) { Set *ichk = 1 if converged. Otherwise, set *ichk = 0. } [in] b[] Array b[lb] (lb >= n) Right hand side vector b. [in,out] x[] Array x[lx] (lx >= n) [in] Initial guess of solution. [out] Obtained approximate solution. [in] maxiter Maximum number of iterations. (maxiter > 0) [out] iter Final number of iterations. [out] res Final residual norm norm(b - A*x). [in] lwork Size of array work[]. (lwork >= 15*n) [out] work[] Array work[lwork] Work array. [out] info = 0: Successful exit < 0: The (-info)-th argument is invalid. = 11: Maximum number of iterations exceeded. = 12, 13, 14, 15: Breakdown occurred.