inst/examples/hmm-univariate-categorical.R

In luisdamiano/BayesHMM: Full Bayesian Inference for Hidden Markov Models

mySpec <- hmm(
  K = 3, R = 1,
  observation = Categorical(
    theta = Dirichlet(alpha = c(1, 1, 1, 1)),
    N = 4
  ),
  initial     = Dirichlet(alpha = c(1, 1, 1)),
  transition  = Dirichlet(alpha = c(1, 1, 1)),
  name = "Univariate Categorical"
)

set.seed(9000)
y = as.matrix(
  c(
    sample(1:4, size = 100, replace = TRUE, prob = c(0.1, 0.1, 0.1, 0.7)),
    sample(1:4, size = 100, replace = TRUE, prob = c(0.1, 0.1, 0.7, 0.1)),
    sample(1:4, size = 100, replace = TRUE, prob = c(0.1, 0.7, 0.1, 0.1))
  )
)

myFit <- fit(mySpec, y = y, chains = 1, iter = 500, seed = 9000)

plot_series(myFit)

print_fit(myFit)