Skip to content

Iblpupil jax backend #52

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
themattinthehatt merged 9 commits into from
Jun 23, 2025
Merged

Iblpupil jax backend #52

Show file tree
Hide file tree
Changes from 8 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
62051f5
quantile variable filter frame length fix
keeminlee Mar 10, 2025
b8dfb1a
Merge branch 'main' of https://github.com/paninski-lab/eks
keeminlee Mar 10, 2025
30b615a
formatting
keeminlee Mar 10, 2025
6abc988
merge
keeminlee Mar 19, 2025
2346357
Merge branch 'main' of https://github.com/paninski-lab/eks
keeminlee Mar 19, 2025
47648b3
Merge branch 'main' of https://github.com/paninski-lab/eks
keeminlee Jun 23, 2025
b94df64
ibl pupil smoother backend now JAX
keeminlee Jun 23, 2025
84a18fc
strange random seeding issue fix
keeminlee Jun 23, 2025
7fc67d7
removal of jax_forward_pass_nlls and redundant funcs
keeminlee Jun 23, 2025
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: 2 additions & 3 deletions eks/core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -390,7 +390,6 @@ def jax_forward_pass(y, m0, cov0, A, Q, C, R, ensemble_vars):
"""
# Initialize carry
carry = (m0, cov0, A, Q, C, 0)

# Run the scan, passing y and ensemble_vars as inputs to kalman_filter_step
carry, outputs = jax.lax.scan(kalman_filter_step, carry, (y, ensemble_vars))
mfs, Vfs, _ = outputs
Expand Down Expand Up @@ -604,8 +603,8 @@ def optimize_smooth_param(
blocks: keypoints to be blocked for correlated noise. Generates on smoothing param per
block, as opposed to per keypoint.
Specified by the form "x1, x2, x3; y1, y2" referring to keypoint indices (start at 0)
maxiter
verbose
maxiter: Maximum iterations before forced optimization loop exit
verbose: Prints extra information for smoothing parameter iterations

Returns:
tuple: Final smoothing parameters, smoothed means, smoothed covariances,
Expand Down
Loading