<div dir="ltr">Thresholding is a heuristic and a crude algorithm (just chop weak edges with the simplest measure of "weak"). I don't know of any abstract analysis of it.<div><br></div><div>The original smoothed aggregation papers by Vanek, Mandel, Brezina were fairly by fairly mathy folks. </div><div><br></div><div>Ray Tuminaro and Luke Olsen have done some nice work to develop a more robust coarsening strategies that do some non-trivial analysis of the matrix to try to identify genuinely strong connections. Crude methods are not bad for M matrices but can get fooled by higher order discretizations.</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Jan 5, 2017 at 9:23 AM, Jeremy Theler <span dir="ltr"><<a href="mailto:jeremy@seamplex.com" target="_blank">jeremy@seamplex.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Yes, I read that page and it was that paragraph that made me want to<br>
learn more.<br>
<br>
For example, that pages says:<br>
<br>
“-pc_gamg_threshold 0.0 is the most robust option (the reason for this<br>
is not obvious) ...”<br>
<br>
<br>
Where can I find more math-based background on this subject? I mean,<br>
some text that describes the methods and not just the implementation as<br>
the source code at gamg/util.c so I can better understand what is going<br>
on.<br>
<br>
<br>
Thanks<br>
<span class="HOEnZb"><font color="#888888"><br>
<br>
<br>
--<br>
Jeremy Theler<br>
<a href="http://www.seamplex.com" rel="noreferrer" target="_blank">www.seamplex.com</a><br>
</font></span><div class="HOEnZb"><div class="h5"><br>
<br>
<br>
On Thu, 2017-01-05 at 09:18 -0500, Mark Adams wrote:<br>
> You want the bottom of page 84 in the manual.<br>
><br>
> On Wed, Jan 4, 2017 at 4:33 PM, Barry Smith <<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>><br>
> wrote:<br>
><br>
> The manual page gives a high-level description<br>
> <a href="http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/PC/PCGAMGSetThreshold.html" rel="noreferrer" target="_blank">http://www.mcs.anl.gov/petsc/<wbr>petsc-current/docs/<wbr>manualpages/PC/<wbr>PCGAMGSetThreshold.html</a> the exact details can be found in the code here <a href="http://www.mcs.anl.gov/petsc/petsc-dev/src/ksp/pc/impls/gamg/util.c.html#PCGAMGFilterGraph" rel="noreferrer" target="_blank">http://www.mcs.anl.gov/petsc/<wbr>petsc-dev/src/ksp/pc/impls/<wbr>gamg/util.c.html#<wbr>PCGAMGFilterGraph</a> I'm adding a link from the former to the later in the documentation.<br>
><br>
> Barry<br>
><br>
><br>
><br>
> > On Jan 4, 2017, at 3:16 PM, Jeremy Theler<br>
> <<a href="mailto:jeremy@seamplex.com">jeremy@seamplex.com</a>> wrote:<br>
> ><br>
> > * Any reference to what pc_gamg_treshold means and/or does?<br>
> ><br>
> ><br>
> ><br>
> > On Wed, 2017-01-04 at 18:13 -0300, Jeremy Theler wrote:<br>
> >> Hi! Any reference to what does -pc_gamg_threshold mean<br>
> and/or?<br>
> >><br>
> ><br>
><br>
><br>
><br>
><br>
<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><i>The secret to doing good research is always be a little underemployed. You waste years by not being able to waste hours</i> -- Amos Tversky<br></div></div></div></div>
</div>