newton

Implements two complementary high-dimensional feature screening
methods, Adaptive Iterative Ridge High-dimensional Ordinary Least-squares
Projection (Air-HOLP, suitable when the number of predictors p is greater
than or equal to the sample size n) and Adaptive Iterative Ridge Ordinary
Least Squares (Air-OLS, for n greater than p). Also provides helper
functions to generate compound-symmetry and AR(1) correlated data, plus
a unified Air() front end and a summary method. For methodological details
see Joudah, Muller and Zhu (2025) <doi:10.1007/s11222-025-10599-6>.

Ibrahim Joudah

AirScreen

Feature Screening via Adaptive Iterative Ridge (Air-HOLP and
Air-OLS)

Samuel Muller

Houying Zhu

newton function

<dl><dt>f</dt>
<dd>Function whose root is sought.</dd>
<dt>fp</dt>
<dd>Derivative function of <code>f</code>.</dd>
<dt>x</dt>
<dd>Numeric starting value.</dd>
<dt>tol</dt>
<dd>Convergence tolerance (default <code>1e-3</code>).</dd>
<dt>m</dt>
<dd>Maximum number of iterations (default \(100\)).</dd></dl>

Arguments

Newton-Raphson root finder — newton

<dl>

<dt>f</dt>
<dd>Function whose root is sought.</dd>


<dt>fp</dt>
<dd>Derivative function of <code>f</code>.</dd>


<dt>x</dt>
<dd>Numeric starting value.</dd>


<dt>tol</dt>
<dd>Convergence tolerance (default <code>1e-3</code>).</dd>


<dt>m</dt>
<dd>Maximum number of iterations (default \(100\)).</dd>

</dl>

newton: Newton-Raphson root finder

Description

Usage

Value

Arguments

Details

Examples