ssm - Fit and analyze state-space models on time-series data
Fits and analyzes HMM, SLDS, and LDS state-space models on time-series data with support for missing data, model comparison, and state extraction.
Tags
Updated: 2026-09-16Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Fit state-space models
- Decode discrete latent states
- Smooth continuous latent trajectories
- Calculate scalar log likelihood
- Interpolate missing time series data
- Sample from generative models
Inputs
- Time-series observation data
- Model configuration parameters
- Data mask arrays
Outputs
- Fitted state-space model objects
- Log-likelihood and ELBO values
- Discrete latent state sequences
- Continuous latent state trajectories
Requirements
- ssm package
- numpy package
- Python environment