probml/dynamax

View on GitHub

A Python package for probabilistic state space modeling with JAX

Stars: 1k
Forks: 114
Open beginner issues: 0
Indexed issues: 47
Avg merge: 19h 14m
Merged PRs (30d): 1

Dominant language: Python
License: MIT
Last GitHub push: Sep 16, 2026
Latest indexed: Sep 20, 2026
Contributing guide: Contributing guide
Code of conduct: No code of conduct
Beginner labels: help wanted

Beginner friendly Everything

Open All

47 open issues indexed

[BUG] progress bar not needed anymore for BlackJAX's MCMC implementation Open

Difficulty 1/5 Under an hour Newbie friendliness 90/100

probml/dynamax#463 · Opened Sep 18, 2026
Interest in differentiable classic infinite horizon LQ control (generating LGSSM types)? Open

Difficulty 5/5 Over a week Newbie friendliness 38/100

probml/dynamax#460 · Opened Aug 14, 2026
Trim heavy dependencies: reconsider hard scikit-learn dependency for KMeans-only usage Open

Difficulty 3/5 1-2 days Newbie friendliness 55/100

probml/dynamax#459 · 1 comment · Opened Aug 4, 2026
Feature request: per-timestep observation uncertainty for Gaussian HMM / ARHMM (heteroscedastic emissions) Open

Difficulty 5/5 Over a week Newbie friendliness 30/100

probml/dynamax#446 · Opened Feb 9, 2026
masked arrays to deal with with missing data Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#430 · 2 comments · 1 reaction · Opened Aug 12, 2025
[Question] Why compute posterior as `prior_cov - K @ S @ K.T`? Open

Difficulty 2/5 1-3 hours Newbie friendliness 35/100

probml/dynamax#422 · Opened Jul 8, 2025
Add check for strictly positive training data for CategoricalHMM Open

Difficulty 3/5 1-2 days Newbie friendliness 35/100

probml/dynamax#421 · Opened Jun 3, 2025
compatibility of dynamax with sts-jax and python Open

Difficulty 4/5 3-5 days Newbie friendliness 30/100

probml/dynamax#420 · 1 reaction · Opened Jun 2, 2025
inputs to ARHMM Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#419 · Opened Jun 2, 2025
[BUG] Incorrect summary statistics in `LinearGaussianConjugateSSM.fit_blocked_gibbs` Open

Difficulty 3/5 1-2 days Newbie friendliness 48/100

probml/dynamax#416 · 3 comments · Opened Apr 18, 2025
NaNs in EM after a few iterations (Cholesky decomposition) Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#404 · 1 comment · Opened Mar 6, 2025
Inhomogeneous Kalman filter Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

probml/dynamax#402 · 1 reaction · Opened Mar 1, 2025
Support missing (`nan`) emissions in `LinearGaussianConjugateSSM.fit_blocked_gibbs` Open

Difficulty 5/5 Over a week Newbie friendliness 30/100

probml/dynamax#399 · 1 comment · Opened Feb 26, 2025
Fix iterated EKS and CMGS Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#398 · Opened Feb 19, 2025
Support for HMMs with num_states=1 Open

Difficulty 3/5 1-2 days Newbie friendliness 35/100

probml/dynamax#380 · 1 comment · Opened Oct 1, 2024
dynmax glm-hmm using jagged input arrays with differing E and M steps Open

Difficulty 5/5 Over a week Newbie friendliness 30/100

probml/dynamax#362 · 4 comments · Opened May 14, 2024
Add general info form message passing algorithms for linear Gaussian SSMs Open

Difficulty 5/5 Over a week Newbie friendliness 30/100

probml/dynamax#361 · Opened Apr 22, 2024
dynamax is weirdly slow in my HMM benchmark Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

probml/dynamax#359 · Opened Feb 22, 2024
NAN returned in large data and large iteration number Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#355 · 1 comment · Opened Jan 26, 2024
Fit CategoricalHMM with available data? Open

Difficulty 3/5 1-2 days Newbie friendliness 25/100

probml/dynamax#345 · 1 comment · Opened Oct 12, 2023
GLM-HMM Open

Difficulty 3/5 1-2 days Newbie friendliness 25/100

probml/dynamax#344 · 4 comments · 7 reactions · Opened Sep 22, 2023
Support for rSLDS? Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

probml/dynamax#338 · 2 comments · 4 reactions · Opened Aug 17, 2023
pass inputs into the LDS model Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#329 · 3 comments · Opened Jul 19, 2023
impose constraints on the transition probability. Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#326 · 1 comment · Opened Jun 20, 2023
Is there a planned implementation of a hierarchical hidden markov model? Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#325 · 1 reaction · Opened May 31, 2023
EM yields NaN after some iterations Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

probml/dynamax#316 · 1 comment · 6 reactions · Opened May 17, 2023
Support for kmeans initialization with vmap Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#315 · 3 comments · Opened May 16, 2023
add support for storing just means or covs during kalman filtering Open

probml/dynamax#291 · 1 assignee · Opened Dec 5, 2022
Add spectral methods for HMM learning Open

enhancement help wanted

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#283 · 2 comments · Opened Nov 15, 2022
implement square root extended Kalman filter Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#279 · Opened Nov 14, 2022
add support for normalizing flows + kalman filters Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#277 · Opened Nov 14, 2022
add block gibbs sampling for HMM learning Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 35/100

probml/dynamax#276 · Opened Nov 14, 2022
add support for parameter learning in NLGSSM models Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 15/100

probml/dynamax#275 · Opened Nov 14, 2022
add support for normalizing flow emission model for HMM Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#274 · Opened Nov 14, 2022
add support for HMM inference with sparse transition matrices Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#273 · 6 comments · Opened Nov 14, 2022
implement unscented particle filter using dynamax and blackjax Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 30/100

probml/dynamax#272 · Opened Nov 14, 2022
implement rao-blackwellised particle filtering using dynamax and blackjax Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#271 · 9 comments · Opened Nov 14, 2022
allow nonlinear model callables to be time-indexed Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#264 · Opened Nov 10, 2022
modify SGD training to avoid using generators Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 38/100

probml/dynamax#256 · 1 comment · Opened Nov 9, 2022
Add one-step prediction method to LGSSM class Open

Difficulty 3/5 1-2 days Newbie friendliness 45/100

probml/dynamax#248 · Opened Nov 3, 2022
Add posterior_sample method to CMGF Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#219 · Opened Oct 5, 2022
check MatrixNormal distribution Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

probml/dynamax#131 · 2 comments · Opened Aug 7, 2022
implement basic particle filtering examples (possibly using blackjax) Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

probml/dynamax#112 · 1 comment · Opened Jul 19, 2022
Make HMM learning work with variable length time series Open

help wanted

probml/dynamax#99 · 12 comments · 1 reaction · 2 assignees · Opened Jul 14, 2022
implement hidden semi-markov models Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 20/100

probml/dynamax#21 · 1 reaction · Opened May 25, 2022
implement non-parametric HMMs Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 25/100

probml/dynamax#20 · 1 reaction · Opened May 25, 2022
implement Laplace-EM algorithm for LGSSM with Poisson likelihoods Open

help wanted

probml/dynamax#13 · 1 assignee · Opened May 23, 2022

probml/dynamax

Get new issues in your inbox