Iterative Lanczos' tau method based on Schur complements.

   a = schursolverpw(x, y, ode, conditions, prec, varargin)

   Piecewiase version. a = schursolverpw(varargin) returns the 
   coefficients a satisfying yn = Pa. Along with the  
   approximation of yn computes an estimation for the error 
   (used as stopping criterium). This version allows for partioned domains.

Input (required)
   x              = independent tau variable (itau object).
   y              = dependent tau variable (dtau oject).
   ode            = ordinary differential equation (operator).
   conditions     = problem conditions (cell of char).
   prec           = required precision to stop.

Input (optional)
   pieces         = number of steps for piecewise approach (integer).
   exact_solution = exact solution.
   step           = step of x vector to show the results.
   apsol          = boolean varargin to show the graphical solution.
   resid          = boolean varargin to show the graphical error.
   coeff          = boolean varargin to show the coefficients a.
   spy            = boolean varargin to show the spy of T matrix.
   maxdim         = maximun dimension of T.
   saves          = name varargin to save the .mat with the results.
   method         = method of schur cmplements.
                    'incr'     = incremental.
                    'incr2'    = incremental second version.
                    'nflu'     = non fixed lu.
                    'iter'     = iterative.

   a              = coefficients such that Pa = yn.

See also
  tau, taupw, tausys and tausyspw.