Skip to content

Latest commit

 

History

History
56 lines (39 loc) · 1.66 KB

TODO.md

File metadata and controls

56 lines (39 loc) · 1.66 KB
Raw

Todo

Misc

  • Inline PLY describing building test problem
    • don't need to load file
    • call TetGen on PLY directly
  • Start working on documentation

Refactoring eik3

  • Rename eik2 to eik2f (or something)
  • Rename eik3 to eik31
  • Move transport-related stuff into a separate module
  • Clean up eik3's API
  • Combine jet3 and hess
  • Simplify logic in update
  • Track t_in and t_out separately... move into separate module along with transport stuff

Improving the eik3 algorithm

  • Get rid of diffraction ftype

Results

  • Ray-traced video of direct arrival wavefront propagating throughout building with opacity selected by amplitude and colored by error
    • coloring by error is tough but doable... one way to go would be to do it automatically would be to use Richardson extrapolation...
  • Fully worked out wedge problem with errors for all quantities

Some notes and observations

As much as possible, we want to get rid of our extra stuff related to updating from edges. All the bde functions... We don't want to set BCs here, since we don't want to do "diffraction" updates.

What's the right way to think about updating? At each step, we want to minimize over the VALID front. Probably the best way to do this is to just do so explicitly. But considering the triangle fan around the newly VALID node makes things a bit simpler.

Pretty sure we want to completely get rid of t_out.

Probably would be good to get rid of do_diff_edge_updates_and_adjust.

Lessons learned

  • Explicit error handling from the get-go

JCP revision

  • factor out caches
  • introduce sfunc