Skip to content

More advanced PDE support in dynestyx - #309

Draft
MatthieuDarcy wants to merge 20 commits into
mainfrom
pde_v2
Draft

More advanced PDE support in dynestyx #309
MatthieuDarcy wants to merge 20 commits into
mainfrom
pde_v2

Conversation

@MatthieuDarcy

@MatthieuDarcy MatthieuDarcy commented Aug 11, 2026

Copy link
Copy Markdown
Contributor

Added a tutorial and some utilities for filtering with PDEs:

  1. Random sensor placements: require interpolation operators
  2. Use of sparse observation matrices (because of the above), Support for sparse observation matrices #298
  3. The use of a GP prior

MatthieuDarcy and others added 20 commits August 7, 2026 12:01
EnKF and now supports sparse observation matrices (PF already did this).

EKF and KF will raise a warning if a sparse matrix is passed and this is likely to break under the hood.

Added a couple tests to verify that computation is the same for both sparse and dense matrices.
KF will now raise an error.

EKF will raise a warning, indicating that the computation is supported but likely inefficient.
Added tests to check warnings/error messages as well verification that computation is identical.
Brings in sparse-matrix support for EnKF/EKF filters (observation
handling, cuthbert/cd_dynamax discrete filters) from
BasisResearch/dynestyx@enkf_sparse_support. No file overlap with
pde_v2's own changes, so the merge applied cleanly.
Fist tutorial illustrating:

1. random sensor placements + interpolation
2. use of sparse observation matrices
3. gp prior
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants