summary.mhmm: Summary method for mixture hidden Markov models

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

Summary method for mixture hidden Markov models

Description

Function summary.mhmm gives a summary of a mixture hidden Markov model.

Usage

## S3 method for class 'mhmm'
summary(
  object,
  parameters = FALSE,
  conditional_se = TRUE,
  log_space = FALSE,
  ...
)

Arguments

object	Mixture hidden Markov model of class mhmm.
parameters	Whether or not to return transition, emission, and initial probabilities. FALSE by default.
conditional_se	Return conditional standard errors of coefficients. See vcov.mhmm for details. TRUE by default.
log_space	Make computations using log-space instead of scaling for greater numerical stability at cost of decreased computational performance. Default is FALSE.
...	Further arguments to vcov.mhmm.

Details

The summary.mhmm function computes features from a mixture hidden Markov model and stores them as a list. A print method prints summaries of these: log-likelihood and BIC, coefficients and standard errors of covariates, means of prior cluster probabilities, and information on most probable clusters.

Value

transition_probs

Transition probabilities. Only returned if parameters = TRUE.

emission_probs

Emission probabilities. Only returned if parameters = TRUE.

initial_probs

Initial state probabilities. Only returned if parameters = TRUE.

logLik

Log-likelihood.

BIC

Bayesian information criterion.

most_probable_cluster

The most probable cluster according to posterior probabilities.

coefficients

Coefficients of covariates.

vcov

Variance-covariance matrix of coefficients.

prior_cluster_probabilities

Prior cluster probabilities (mixing proportions) given the covariates.

posterior_cluster_probabilities

Posterior cluster membership probabilities.

classification_table

Cluster probabilities (columns) by the most probable cluster (rows).

See Also

build_mhmm and fit_model for building and fitting mixture hidden Markov models; and mhmm_biofam for information on the model used in examples.

Examples

# Loading mixture hidden Markov model (mhmm object)
# of the biofam data
data("mhmm_biofam")

# Model summary
summary(mhmm_biofam)

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