<div class="gmail_quote">On Thu, Feb 2, 2012 at 11:14, Klaij, Christiaan <span dir="ltr"><<a href="mailto:C.Klaij@marin.nl">C.Klaij@marin.nl</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div id=":1qq"> Mat A,subA[4];<br>
<br>
MatCreate(PETSC_COMM_WORLD,&subA[0]);<br>
MatSetSizes(subA[0],PETSC_DECIDE,PETSC_DECIDE,2*nx*ny,2*nx*ny);<br>
/* set type, set prefix, set values here */<br>
<br>
MatCreate(PETSC_COMM_WORLD,&subA[1]);<br>
MatSetSizes(subA[1],PETSC_DECIDE,PETSC_DECIDE,2*nx*ny,nx*ny);<br>
/* set type, set prefix, set values */<br>
<br>
MatTranspose(subA[1],MAT_INITIAL_MATRIX,&subA[2]);<br>
<br>
MatCreate(PETSC_COMM_WORLD,&subA[3]);<br>
MatSetSizes(subA[3],PETSC_DECIDE,PETSC_DECIDE,nx*ny,nx*ny);<br>
/* set type, set prefix, set values */<br>
<br>
MatCreateNest(PETSC_COMM_WORLD,2,PETSC_NULL,2,PETSC_NULL,subA,&A);<br></div></blockquote><div><br>I should add a MatNestGetISs() so you can get out the automatically-created ISs. They will have the structure described below. If you want to work with the released version, you should create the ISs yourself.<br>
</div><blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div id=":1qq">
<div class="im"><br>
<br>
> This really can't be right. Modify ex28 to have it print out the index<br>
> sets. If distributed evenly over two procs, we might expect the first index<br>
> set to hold [0..11; 18..29] and the second to hold [12..17; 30..35]. The<br>
> way you are addressing would force a non-contiguous row partition.<br>
<br>
</div>The partitioning corresponds to the grid. With the 3x4 grid and 2<br>
procs, I have 6 cells per proc. subA[0] corresponds to two<br>
variables per cell (u and v) so 12 rows on proc0 and 12 on<br>
proc1. SubA[3] has one variable per cell (p) so 6 rows on proc0<br>
and 6 on proc1. I tried to create the matching ISs but something's wrong.<br>
What would be the right way to do it?</div></blockquote></div><br>A has a contigous distribution, so the ISs must respect that. Did you try creating the index sets described above? Please explain "something's wrong".<br>