[Users] 8th vs 6th order

[Erik Schnetter]

On Oct 22, 2025, at 11:51, Rhiannon Silva <SilvaRL at cardiff.ac.uk> wrote:
> 
> Hi, 
> 
> The GW150914 example parameter file uses 8th order finite differencing and I need a 6th order version for comparison in terms of both accuracy and speed. Apart from changing the finite differencing orders from 8 to 6, and I assume GlobalDerivative automatically selects the correct dissipation order since it's not specified, what would I use for the following values?
> 
> Driver::ghost_size                      = 5
> Coordinates::patch_boundary_size        = 5
> Coordinates::additional_overlap_size    = 3
> Coordinates::outer_boundary_size        = 5

You would set "ghost_size = 4" and "patch_boundary_size = 4" and "outer_boundary_size = 4". You might be able to reduce "additional_overlap_size" as well.

Carpet::prolongation_order_space        = 5

It is up to you whether you reduce "prolongation_order_space". Probably not.

-erik

> Is there anything else I missed that I should change?





-- 
Erik Schnetter <eschnetter at perimeterinstitute.ca>
http://www.perimeterinstitute.ca/personal/eschnetter/