<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<font face="Ubuntu">Hi Jed,<br>
<br>
thank you for your reply.<br>
I checked the PETSc version currently installed on my system (Blue
Gene P); it is 3.1-p2<br>
so I will use the DA type.<br>
Also, I cannot take advantage of DMDAGlobalToNatural since version
3.1 does not support it.<br>
I was able to solve a simple Poisson system successfully but the
solution vector is not in the correct order<br>
and I do not understand how to fix this. Is there an equivalent to
</font><font face="Ubuntu">DMDAGlobalToNatural in version 3.1?<br>
<br>
Thank you,<br>
Michele<br>
</font><font face="Ubuntu"><br>
<br>
</font>
<div class="moz-cite-prefix">On 06/26/2012 11:36 AM, Jed Brown
wrote:<br>
</div>
<blockquote
cite="mid:CAM9tzSk_97JUKhiSTXF_zP1MQKeGEFt901dC3w=t962szym8rA@mail.gmail.com"
type="cite">
<div class="gmail_quote">On Tue, Jun 26, 2012 at 10:24 AM, Michele
Rosso <span dir="ltr"><<a moz-do-not-send="true"
href="mailto:mrosso@uci.edu" target="_blank">mrosso@uci.edu</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"> <font face="Ubuntu">Thanks
Jed for your reply.<br>
<br>
1) I will use the DMDA object type then. I am still not
very clear about the difference between DA and DMDA
though.<br>
</font></div>
</blockquote>
<div><br>
</div>
<div>DM is the generic interface, DMDA is the structured grid
implementation. It used to be called just DA and managed a bit
different. Maybe you were looking at old docs?</div>
<div> </div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"><font face="Ubuntu"> <br>
2) I am not interested in having ghost nodes updated. I
want the proper values of the solution on the proper
processor.<br>
I have to fill the known-terms-vector with
nodes-dependent values ( in contrast with </font>ex45f.F,
where vector b is filled with 1s, thus there is<br>
no dependence on the grid location). Since every
processor defines "non-local" (according to the PETSc
internal ordering) components, the vector is re-arranged<br>
and so is the solution vector. So I will have on every
process a solution which partially should be on a
difference process. And this is not about ghost cell.<br>
Sorry for this long explanation but I am trying to be as
clear as I can.<br>
</div>
</blockquote>
<div><br>
</div>
<div><a moz-do-not-send="true"
href="http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/DM/DMDAGlobalToNaturalBegin.html">http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/DM/DMDAGlobalToNaturalBegin.html</a></div>
<div> </div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"> <br>
Thank you fro your help and patience,<br>
<br>
Michele
<div>
<div class="h5"><br>
<br>
<font face="Ubuntu"><br>
<br>
<br>
<br>
<br>
<br>
<br>
</font>
<div>On 06/26/2012 10:42 AM, Jed Brown wrote:<br>
</div>
<blockquote type="cite">
<div class="gmail_quote">On Tue, Jun 26, 2012 at 9:37
AM, Michele Rosso <span dir="ltr"><<a
moz-do-not-send="true"
href="mailto:mrosso@uci.edu" target="_blank">mrosso@uci.edu</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"> <font
face="Ubuntu">Hi,<br>
<br>
I need some help to use the PETSc library
inside my Fortran 95 code.<br>
My goal is to solve a symmetric linear system
that derives from the finite difference
discretization<br>
of the Poisson equation. I will use the
preconditioned conjugate method.<br>
<br>
I am mostly interested in how to decompose my
data among the different processes. <br>
In particular:<br>
<br>
1) Since my code already implements the
2D-decomposition, would it be best to build
the matrix with the DMDA object type, DA
object type<br>
or the regular Mat type? <br>
</font></div>
</blockquote>
<div><br>
</div>
<div>It is certainly easiest to just use DMDA (and
you will get geometric multigrid for free, which
is unbeatable for this problem), but you can work
directly with Mat if you prefer. See
src/ksp/ksp/examples/tutorials/ex45f.F for a
Fortran example.</div>
<div> </div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"><font
face="Ubuntu"> <br>
2) After inserting values into a
vector/matrix, PETSc performs any needed
message passing of nonlocal components, thus
the values locally contained on a process may
be communicated to another process. How can I
revert this at the end of the computation,
that is, how can I be sure that the local
solution vector contains the values associated
to the grid nodes contained into the hosting
process?<br>
</font></div>
</blockquote>
<div><br>
</div>
<div>Please read the section of the user's manual on
local versus global spaces and the structured grid
decompositions. If you use DM, there is
DMGlobalToLocalBegin/End that update the ghost
points in the local vectors.</div>
<div> </div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"><font
face="Ubuntu"> <br>
<br>
Thank you,<br>
<br>
Michele</font> </div>
</blockquote>
</div>
<br>
</blockquote>
<br>
<br>
</div>
</div>
</div>
</blockquote>
</div>
<br>
</blockquote>
<br>
<br>
</body>
</html>