fit_sequence_hmm_mixture: Fit a mixture of categorical hidden Markov models

In gp3sequences: Transparent Analysis of Ordered Categorical Sequences

Fit a mixture of categorical hidden Markov models

Description

Fits sequence-level mixture components, each containing a categorical HMM. Component membership is a statistical clustering device and should not be interpreted as a substantive latent type without external validation.

Usage

fit_sequence_hmm_mixture(
  data,
  n_components,
  n_states,
  sequence_id_col = "sequence_id",
  order_col = "sequence_order",
  state_col = "state",
  symbol_levels = NULL,
  max_iter = 200L,
  inner_initial_iter = 20L,
  tolerance = 1e-06,
  pseudocount = 1e-06,
  seed = 1L
)

Arguments

data	Long-format sequence data.
n_components	Number of mixture components.
n_states	Number of hidden states per component; scalar or vector.
sequence_id_col, order_col, state_col	Sequence columns.
symbol_levels	Optional symbol ordering.
max_iter	Maximum mixture-EM iterations.
inner_initial_iter	Initial single-HMM iterations per component.
tolerance	Relative log-likelihood tolerance.
pseudocount	Smoothing count.
seed	Reproducibility seed.

Value

An object of class gp3_sequence_hmm_mixture.

Examples

sequences <- data.frame(
  sequence_id = rep(c("s1", "s2", "s3", "s4"), each = 4L),
  sequence_order = rep(1:4, times = 4L),
  state = c("A", "B", "C", "D", "A", "B", "C", "C",
            "D", "C", "B", "A", "D", "C", "A", "A"),
  group = rep(c("g1", "g2"), each = 8L),
  stringsAsFactors = FALSE
)
fit_sequence_hmm_mixture(sequences, n_components = 2L, n_states = 2L,
                         max_iter = 5L, inner_initial_iter = 2L, seed = 1L)

gp3sequences documentation built on Aug. 23, 2026, 5:10 p.m.