[Carpet] ApplyBCs and global mode scheduling

[Erik Schnetter]

On Oct 19, 2006, at 05:16:19, Bela Szilagyi wrote:

> It seems that the two "types" of scheduling (i.e., "global" and  
> "level") are
> introducing all kinds of subtle inconsistencies in the Carpet  
> user's code.
> Some of these may be fixed by editing the local schedule.ccl files,  
> but
> others are a harder nut.
>
> Take, e.g., a thorn of mine which wants to convert from the "harmonic"
> variables into the ADM variables at post-step, before  
> AHFinderDirect finds
> its horizons.

PostStep is probably not the right place to convert to the ADM  
variables.  I would either schedule it "IN MoL_PostStep", or "AT evol  
AFTER MoL_Evolution".  This guarantees also that the ADM variables  
are restricted, which keeps them consistent.  You should calculate  
them without any schedule options.

If you do that, then everything will be automatically consistent.

Applying the boundary conditions in the way you describe will perform  
unnecessary work, since it will convert and apply the boundary  
conditions of the coarse grid at each fine grid time step.



In principle, you are right that the scheduling options are not  
always very convenient to use.  In this case, I suggest to write a  
scheduling script.  This is a C++ routine which you schedule, and  
which calls (schedules) other routines or groups in a very specific  
order.  This is often much easier than writing a complex schedule.   
You can also loop over levels there.  Without thinking too deeply, I  
think that scheduling ApplyBCs from there would do the correct  
thing.  Let me know if you need an example.

-erik

-- 
Erik Schnetter <schnetter at cct.lsu.edu>

My email is as private as my paper mail.  I therefore support encrypting
and signing email messages.  Get my PGP key from www.keyserver.net.



-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : /archives/developers/attachments/20061020/0181b3ee/attachment.pgp