simulate_pars: Simulate Parameters of Hidden Markov Models

In seqHMM: Mixture Hidden Markov Models for Social Sequence Data and Other Multivariate, Multichannel Categorical Time Series

Simulate Parameters of Hidden Markov Models

Description

These are helper functions for quick construction of initial values for various model building functions. Mostly useful for global optimization algorithms which do not depend on initial values.

Usage

simulate_initial_probs(n_states, n_clusters = 1)

simulate_transition_probs(
  n_states,
  n_clusters = 1,
  left_right = FALSE,
  diag_c = 0
)

simulate_emission_probs(n_states, n_symbols, n_clusters = 1)

Arguments

n_states	Number of states in each cluster.
n_clusters	Number of clusters.
left_right	Constrain the transition probabilities to upper triangular. Default is FALSE.
diag_c	A constant value to be added to diagonal of transition matrices before scaling.
n_symbols	Number of distinct symbols in each channel.

See Also

build_hmm, build_mhmm, build_mm, build_mmm, and build_lcm for constructing different types of models.

seqHMM documentation built on July 9, 2023, 6:35 p.m.