Skip to content

Add regularisation term before Cholesky decomposition #459

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Apr 17, 2025
+5 −0
Merged

Add regularisation term before Cholesky decomposition #459

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions pymc_extras/inference/pathfinder/pathfinder.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -489,6 +489,7 @@ def bfgs_sample_dense(

N = x.shape[-1]
IdN = pt.eye(N)[None, ...]
IdN += IdN * REGULARISATION_TERM

# inverse Hessian
H_inv = (
Expand Down Expand Up @@ -568,9 +569,13 @@ def bfgs_sample_sparse(
# qr_input: (L, N, 2J)
qr_input = inv_sqrt_alpha_diag @ beta
(Q, R), _ = pytensor.scan(fn=pt.nlinalg.qr, sequences=[qr_input], allow_gc=False)

IdN = pt.eye(R.shape[1])[None, ...]
IdN += IdN * REGULARISATION_TERM

Lchol_input = IdN + R @ gamma @ pt.matrix_transpose(R)

# TODO: make robust Lchol calcs more robust, ie. try exceptions, increase REGULARISATION_TERM if non-finite exists
Lchol = pt.linalg.cholesky(Lchol_input, lower=False, check_finite=False, on_error="nan")

logdet = 2.0 * pt.sum(pt.log(pt.abs(pt.diagonal(Lchol, axis1=-2, axis2=-1))), axis=-1)
Expand Down
Loading