VarReg.control

Use <code>VarReg.control</code> to determine parameters for the fitting of <code>semiVarReg</code>. Typically only used internally within functions.

Methods for fitting semi-parametric mean and variance models, with normal or censored data. Extended to allow a regression in the location, scale and shape parameters, and further for multiple regression in each.

Kristy Robledo

VarReg

Semi-Parametric Variance Regression

Ian Marschner

VarReg.control function

<dl><dt>bound.tol</dt>
<dd>Positive tolerance for specifying the interior of the parameter space. This allows the algorithm to terminate early if an interior maximum is found. If set to <code>bound.tol=Inf</code>, no early termination is attempted.</dd>
<dt>epsilon</dt>
<dd>Positive convergence tolerance. If \(\theta\) is a vector of estimates, convergence is declared when \(\sqrt{(\sum (\theta_{old} - \theta_{new})^2)}/ \sqrt{\sum (\theta_{old})^2} \). This should be smaller than <code>bound.tol</code>.</dd>
<dt>maxit</dt>
<dd>integer giving the maximum number of EM algorithm iterations for a given parameterisation.</dd></dl>

Arguments

Auxillary for controlling VarReg fitting — VarReg.control

<dl>

<dt>bound.tol</dt>
<dd>Positive tolerance for specifying the interior of the parameter space. This allows the algorithm to terminate early if an interior maximum is found. If set to <code>bound.tol=Inf</code>, no early termination is attempted.</dd>


<dt>epsilon</dt>
<dd>Positive convergence tolerance. If \(\theta\) is a vector of estimates, convergence is declared when \(\sqrt{(\sum (\theta_{old} - \theta_{new})^2)}/ \sqrt{\sum (\theta_{old})^2} \). This should be smaller than <code>bound.tol</code>.</dd>


<dt>maxit</dt>
<dd>integer giving the maximum number of EM algorithm iterations for a given parameterisation.</dd>

</dl>

VarReg.control: Auxillary for controlling VarReg fitting

Description

Usage

Value

Arguments

Details