[ET Trac] [Einstein Toolkit] #1132: CarpetEvolutionMask: Take horizon into account

[Einstein Toolkit]

#1132: CarpetEvolutionMask: Take horizon into account
--------------------------+-------------------------------------------------
  Reporter:  eschnett     |       Owner:     
      Type:  enhancement  |      Status:  new
  Priority:  major        |   Milestone:     
 Component:  Other        |     Version:     
Resolution:               |    Keywords:     
--------------------------+-------------------------------------------------

Comment (by rhaas):

 I am not sure if EvolutionMask is the right place for this. Right now it
 tracks which points will never influence the result of a
 reduction/interpolation output and thus need not be evolved at all. The
 horizon does not fall into this category I think since the inside of the
 horizon is "unphysical and should be used" rather than "no value will
 every be used for output". Ie. NaNs inside the horizon will propagate
 outwards but NaNs at points where EvolutionMask is zero (non-evolved
 points) will not propagate.

 Please note that AHFinderDirect already offers (two) ways of setting a
 mask when points are inside the horizon. Its emask grid function and
 spacemask support. For hydro there is also Hydro_SetMask_SphericalSurface
 which can set GRHydro's excision mask to any spherical surface (eg. set by
 AHFinderDirect) and which causes GRhydro to not evolve these points
 (internally the atmosphere_mask is set and points are forced to be
 atmosphere).

-- 
Ticket URL: <https://trac.einsteintoolkit.org/ticket/1132#comment:1>
Einstein Toolkit <http://einsteintoolkit.org>
The Einstein Toolkit