<div class="gmail_quote">On Thu, Dec 29, 2011 at 11:51, Dave Nystrom <span dir="ltr"><<a href="mailto:Dave.Nystrom@tachyonlogic.com">Dave.Nystrom@tachyonlogic.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div id=":780">Both although the most challenging problem right now is one for a Hall matrix<br>
system that is block banded with 2x2 blocks.</div></blockquote><div><br></div><div>What is the null space of this operator.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div id=":780"> Right now, I am just inputting<br>
the matrix into a petsc seqaij matrix. I guess I need to put it into a petsc<br>
block matrix but have not yet attempted to do that. I'm not sure how much<br>
work is involved - I need to read the section in the users manual on that.<br>
<br>
Anyway, I have 5 scalar systems and on vector system with 2x2 blocks. I am<br>
interested in applying gamg to all of them to see how well it works.</div></blockquote></div><br><div>PCGAMG currently takes coordinates for a nodal basis, from which it generates a null space (making assumptions about the equations). I have it on my list to update so that you can call MatSetNearNullSpace() and have PCGAMG use this null space.</div>