lc: Fit latent-class MPT models

In methexp/HMMTreeR: An R Interface to HMMTree

Description

This function estimates one or more latent-class MPT models. Starting from a one-class solution, it increments the number of latent classes until a criterion is met.

Usage

lc(
  model_file,
  data_file,
  classes = NULL,
  max_classes = 20,
  runs = 20,
  fisher_information = "expected",
  montecarlo_samples = 1e+05,
  crit = "AIC",
  verbose = FALSE,
  keep_files = FALSE
)

Arguments

model_file	Character. File path to an .eqn file
data_file	Character. File path to a data file
classes, max_classes	The number of classes for a single model or the maximum number of classes. To estimate a single model, specify classes; to estimate multiple models with increasing number of classes, specify max_classes.
runs	Integer. Number of optimization runs.
fisher_information	The type of Fisher Information to be computed. Can be either expected, montecarlo, observed, or none. Defaults to expected. However, if expected Fisher Information cannot be computed, the Monte Carlo method is used.
montecarlo_samples	The number of simulations to be used for Monte Carlo Fisher Information. Defaults to 1e5.
crit	If multiple models with an increasing number of classes are to be estimated, use this criterion to stop estimating more complex models. Can be one of "AIC", "BIC", or "BIC2".
verbose	Logical. Indicating whether an announcement is printed on the console when fitting of a new model starts.
keep_files	Logical. Should temporary files be retained?

References

Stahl, C., & Klauer, K.C. (2007). HMMTree: A computer program for hierarchical multinomial processing tree models. Behavior Research Methods, 39, 267-273.

methexp/HMMTreeR documentation built on Nov. 11, 2021, 7:31 p.m.