On 7/24/19 12:23 PM, 'Maxi Miller' via deal.II User Group wrote:
> My LinearOperator only provides a vmult-interface, nothing else, but for 
> initializing the preconditioner I still need a sparse matrix, thus I can not 
> use the LinearOperator, as far as I understand. Or is there a way to use it?

You just found the problem everyone faces. The LinearOperator interface is 
useful if you only know how to express a linear operator through its action, 
without knowing the entries of the matrix that represent this action. This is 
all you need for the iterative solvers we have.

But to work well, these iterators all need preconditioners, and most of the 
preconditioners we have require knowledge of the entries of the matrix -- 
i.e., there is no longer any advantage to using LinearOperator over building a 
matrix. The ways you can work around this are Chebyshev iterations, for 
example; step-20 also in great detail explores this issue. But the point is 
that it's a difficult issue.

Best
  W.


-- 
------------------------------------------------------------------------
Wolfgang Bangerth          email:                 bange...@colostate.edu
                            www: http://www.math.colostate.edu/~bangerth/

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/dealii/5f59e9e1-ff5d-d1da-9671-bf18facfb595%40colostate.edu.

Reply via email to