gradient

Calculates the gradient function of a <code>model</code> with respect to the parameters, <code>char_vars</code>, evaluates
it at the provided <code>values</code> and returns the result as a function of the variable <code>x</code>.

Calculates D-, Ds-, A-, I- and L-optimal designs for non-linear models, via an implementation of the cocktail algorithm (Yu, 2011, <doi:10.1007/s11222-010-9183-2>). Compares designs via their efficiency, and augments any design with a controlled efficiency. An efficient rounding function has been provided to transform approximate designs to exact designs.

Carlos de la Calle-Arroyo

optedr

Calculating Optimal and D-Augmented Designs

Jesús López-Fidalgo

Licesio J. Rodríguez-Aragón

gradient function

<dl><dt>model</dt>
<dd>formula describing the model, which must contain only <code>x</code>, the parameters defined in
<code>char_vars</code> and the numerical operators.</dd>
<dt>char_vars</dt>
<dd>character vector of the parameters of the model.</dd>
<dt>values</dt>
<dd>numeric vector with the nominal values of the parameters in <code>char_vars</code>.</dd>
<dt>weight_fun</dt>
<dd>optional function variable that represents the square of the structure of variance, in case of heteroscedastic variance of the response</dd></dl>

Arguments

Gradient function — gradient

<dl>

<dt>model</dt>
<dd>formula describing the model, which must contain only <code>x</code>, the parameters defined in
<code>char_vars</code> and the numerical operators.</dd>


<dt>char_vars</dt>
<dd>character vector of the parameters of the model.</dd>


<dt>values</dt>
<dd>numeric vector with the nominal values of the parameters in <code>char_vars</code>.</dd>


<dt>weight_fun</dt>
<dd>optional function variable that represents the square of the structure of variance, in case of heteroscedastic variance of the response</dd>

</dl>

gradient: Gradient function

Description

Usage

Value

Arguments