bootstrap_gmmsslm: Bootstrap Analysis for gmmsslm

In gmmsslm: Semi-Supervised Gaussian Mixture Model with a Missing-Data Mechanism

Bootstrap Analysis for gmmsslm

Description

This file provides functions to perform bootstrap analysis on the results of the gmmsslm function.

This function performs non-parametric bootstrap to assess the variability of the gmmsslm function outputs.

Usage

bootstrap_gmmsslm(
  dat,
  zm,
  pi,
  mu,
  sigma,
  paralist,
  xi,
  type,
  iter.max = 500,
  eval.max = 500,
  rel.tol = 1e-15,
  sing.tol = 1e-15,
  B = 2000
)

Arguments

dat	A matrix where each row represents an individual observation.
zm	A matrix or data frame of labels corresponding to dat.
pi	A numeric vector representing the mixing proportions.
mu	A matrix representing the location parameters.
sigma	An array representing the covariance matrix or list of covariance matrices.
paralist	A list of parameters.
xi	A numeric value representing the coefficient for a logistic function of the Shannon entropy.
type	A character value indicating the type of Gaussian mixture model.
iter.max	An integer indicating the maximum number of iterations.
eval.max	An integer indicating the maximum number of evaluations.
rel.tol	A numeric value indicating the relative tolerance.
sing.tol	A numeric value indicating the singularity tolerance.
B	An integer indicating the number of bootstrap samples.

Value

A list containing mean and sd of bootstrap samples for pi, mu, sigma, and xi.

gmmsslm documentation built on June 8, 2025, 2 p.m.