loop_lss: The EM loop for the LSS model

In VarReg: Semi-Parametric Variance Regression

The EM loop for the LSS model

Description

loop_lss is the EM loop function for the LSS model to be utilised by various other higher level functions

Usage

loop_lss(
  alldat,
  xiold,
  omega2old,
  nuold,
  mean.ind,
  var.ind,
  nu.ind,
  para.space,
  maxit,
  eps,
  int.maxit,
  print.it
)

Arguments

alldat	Dataframe containing all the data for the models. Outcome in the first column.
xiold	Vector of initial location parameter estimates to be fit in the location model.
omega2old	Vector of initial scale2 parameter estimates to be fit in the scale2 model.
nuold	Vector of initial nu parameter estimates to be fit in the nu model.
mean.ind	Vector containing the column numbers of the data in 'alldat' to be fit as covariates in the location model.
var.ind	Vector containing the column numbers of the data in 'alldat' to be fit as covariates in the scale2 model. FALSE indicates a constant variance model.
nu.ind	Vector containing the column numbers of the data in 'alldat' to be fit as covariates in the nu model. NULL indicates constant model.
para.space	Parameter space to search for variance parameter estimates. "positive" means only search positive parameter space, "negative" means search only negative parameter space and "all" means search all.
maxit	Number of maximum iterations for the main EM algorithm.
eps	Very small number for the convergence criteria.
int.maxit	Number of maximum iterations for the internal EM algorithm for the location and scale.
print.it	Logical to indicate if the estimates for each iteration should be printed.

Value

A list of the results from the algorithm, including conv, reldiff, it, mean, xi.new, omega2.new, nu.new, fitted.xi

• conv: Logical argument indicating if convergence occurred

• it: Total iterations performed of the EM algorithm

• reldiff: the positive convergence tolerance that occured at the final iteration

• xinew: Vector of location parameter estimates

• omega2new: Vector of scale squared parameter estimates

• nunew: Vector of shape parameter estimates

• fitted.xi: Vector of fitted location estimates

VarReg documentation built on May 31, 2023, 8:44 p.m.