You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hey @yaugenst-flex can you take on this project after slab_bounds? I think it will be pretty natural after that issue and will get you more exposure to the wild world of tidy3d adjoint. Also was requested by a customer so will be a high impact feature. Thanks!
Basically the user may have some vertices that define a JaxPolySlab but become crossing under erosion / dilation / or sidewall extrusion. We currently use the ComplexPolySlab plugin to handle such cases, which lets the user split these crossing poly slabs into several individual ones. But this is not compatible with JaxPolySlab.
I think a simple test case can be a triangle PolySlab with several vertices and a sidewall_angle > 0. By adjusting the number of vertices and the sidewall_angle, we can trigger this validator to fail. A new JaxComplexPolySlab wrapper should allow us to split it into JaxPolySlab and let the user select one (or a few) to model in their JaxStructure.
@yaugenst-flex I take back what I said earlier about the ordering. This one might be a better introduction to #1117 because it won't really involve any gradient checking. Just need to make sure that we can create the JaxPolySlabs and the gradient information is retained. I might recommend you start with this issue as a JaxPolySlab intro
Hey @yaugenst-flex can you take on this project after
slab_bounds
? I think it will be pretty natural after that issue and will get you more exposure to the wild world of tidy3dadjoint
. Also was requested by a customer so will be a high impact feature. Thanks!Basically the user may have some
vertices
that define aJaxPolySlab
but become crossing under erosion / dilation / or sidewall extrusion. We currently use theComplexPolySlab
plugin to handle such cases, which lets the user split these crossing poly slabs into several individual ones. But this is not compatible withJaxPolySlab
.I think a simple test case can be a triangle
PolySlab
with several vertices and asidewall_angle > 0
. By adjusting the number of vertices and the sidewall_angle, we can trigger this validator to fail. A newJaxComplexPolySlab
wrapper should allow us to split it intoJaxPolySlab
and let the user select one (or a few) to model in theirJaxStructure
.Here's a notebook showing how complex polyslab works: https://docs.flexcompute.com/projects/tidy3d/en/latest/api/_autosummary/tidy3d.plugins.polyslab.ComplexPolySlab.html
After this one, I think we can start digging into a bit more long term projects, and architecture work with some minor improvements sprinkled in.
Thanks a lot!
The text was updated successfully, but these errors were encountered: