miss.glm.control: Auxiliary for Controlling Fitting

In misaem: Linear Regression and Logistic Regression with Missing Covariates

Auxiliary for Controlling Fitting

Description

Auxiliary function for miss.glm fitting. Typically only used internally by miss.glm.fit.

Usage

miss.glm.control(
  maxruns = 500,
  tol_em = 1e-07,
  nmcmc = 2,
  tau = 1,
  k1 = 50,
  subsets = NA,
  seed = NA,
  print_iter = TRUE,
  var_cal = TRUE,
  ll_obs_cal = TRUE,
  alpha = 0,
  lambda = 0,
  save_trace = FALSE,
  init_method = "mean",
  ignore_errors = FALSE
)

Arguments

maxruns	maximum number of iterations. The default is maxruns = 500.
tol_em	the tolerance to stop SAEM. The default is tol_em = 1e-7.
nmcmc	the MCMC length. The default is nmcmc = 2.
tau	rate \tau in the step size (k-k_{1})^{-\tau}. The default is tau = 1.
k1	number of first iterations k_{1} in the step size (k-k_{1})^{-\tau}. The default is k1=50.
subsets	Index of selected covariates if any. The default is all the covariates.
seed	an integer as a seed set for the random generator.
print_iter	logical indicating if output should be produced for each iteration.
var_cal	logical indicating if the variance of estimated parameters should be calculated.
ll_obs_cal	logical indicating if the observed log-likelihood should be calculated.
alpha	elastic-net mixing parameter, with 0 <= alpha <= 1. The default is alpha = 0 (ridge).
lambda	non-negative regularization parameter. The default is lambda = 0 (no regularization).
save_trace	logical indicating if the trace of beta values over iterations should be saved.
init_method	method to initialize the missing data. "mean", its default value, is the only option currently implemented.
ignore_errors	logical indicating if errors during the fitting process should be ignored (TRUE, raising a warning instead) or stop the execution (FALSE). Default is FALSE.

Value

A list with components named as the arguments.

Examples

## For examples see example(miss.glm)

misaem documentation built on Sept. 12, 2025, 1:08 a.m.