TheseqHMM package is designed for fitting hidden(latent) Markov models (HMMs) and their variations for social sequencedata and other categorical sequence data (e.g. categorical time seriesand panel data). Restricted and extended variants include mixture HMMs,Markov models and their mixtures, latent class models, non-homogeneoushidden Markov models (NHMMs) and their mixtures, and feedback-augmentedhidden Markov models (FAN-HMMs).
The package supports models for one or multiple subjects with one ormultiple parallel outcome sequences (channels). External covariates canbe added to explain cluster membership in mixture models, and NHMMs andtheir variants support covariates in initial, transition and emissionmatrices as well.
Maximum likelihood estimation via EM algorithm and direct numericalmaximization with analytical gradients is supported. All main algorithmsare written in C++. Parallel computation is available via OpenMP(pre-2.0.0 version models) andfuture (viaparallel multistart estimation with random initial values).
When using the package in publications, please cite:
Helske, Satu and Helske, Jouni (2019). Mixture hidden Markov modelsfor sequence data: the seqHMM package in R.Journal of StatisticalSoftware, 88(3).doi:10.18637/jss.v088.i03.
Helske, Jouni (2025). Feedback-augmented Non-homogeneous HiddenMarkov Models for Longitudinal Causal Inference.arXivpreprint.doi:10.48550/arXiv.2503.16014.
If you find bugs, please add a new issue here in GitHub. You can alsocontact Satu Helske (firstname.lastname@utu.fi) or Jouni Helske(firstname.lastname@iki.fi). We would be happy to hear yourfeedback.
The package is available on CRAN. Install it via
install.packages("seqHMM")If you want to try the development version of theseqHMMpackage, install it from theR-universe:
install.packages("seqHMM",repos ="https://helske.r-universe.dev")