search

spinsphotonics / fdtdz

Fast, scalable, accessible photonic simulation
MIT License
110 stars 13 forks source link

wrong shape constraint on PML parameters? #27

Closed flaport closed 9 months ago

flaport commented 9 months ago

Hi! It's me again 🙂

Maybe I'm a little bit confused, but currently fdtdz_jax seems to require the pml parameters pml_kappa, pml_sigma and pml_alpha to be of shape 128-sum(pml_widths). is this correct? I would expect these pml parameters to extend all the way to the edges of the z-space.

flaport commented 9 months ago

Or is it because a PML is more memory intensive that the z-span needs to be reduced further?

JesseLu commented 9 months ago

No this is correct. Each layer of pml cost an additional layer of overhead in z.

On Wed, Dec 6, 2023, 2:31 PM Floris Laporte @.***> wrote:

Or is it because a PML is more memory intensive that the z-span needs to be reduced further?

— Reply to this email directly, view it on GitHub https://github.com/spinsphotonics/fdtdz/issues/27#issuecomment-1843793388, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFQRGNLILGFZU2ORV4OLWTYIDW4TAVCNFSM6AAAAABAKDHOBKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNBTG44TGMZYHA . You are receiving this because you are subscribed to this thread.Message ID: @.***>

flaport commented 9 months ago

Thanks!