.smoothElasticNetHessian

smoothed version of non-differentiable elastic LASSO Hessian

internal

Provides regularized structural equation modeling
(regularized SEM) with non-smooth penalty functions (e.g., lasso) building
on 'lavaan'. The package is heavily inspired by the
['regsem'](<https://github.com/Rjacobucci/regsem>) and
['lslx'](<https://github.com/psyphh/lslx>) packages.

Jannik H. Orzek

lessSEM

Non-Smooth Regularization for Structural Equation Models

.smoothElasticNetHessian function

<dl><dt>parameters</dt>
<dd>vector with labeled parameter values</dd>
<dt>tuningParameters</dt>
<dd>list with fields lambda (tuning parameter value), alpha (0&lt;alpha&lt;1. Controls the weight of ridge and lasso terms. alpha = 1 is lasso, alpha = 0 ridge)</dd>
<dt>penaltyFunctionArguments</dt>
<dd>list with fields regularizedParameterLabels (labels of regularized parameters), and eps (controls the smooth approximation of non-differential penalty functions (e.g., lasso, adaptive lasso, or elastic net). Smaller values result in closer approximation, but may also cause larger issues in optimization.)</dd></dl>

Arguments

.smoothElasticNetHessian — .smoothElasticNetHessian

<dl>

<dt>parameters</dt>
<dd>vector with labeled parameter values</dd>


<dt>tuningParameters</dt>
<dd>list with fields lambda (tuning parameter value), alpha (0&lt;alpha&lt;1. Controls the weight of ridge and lasso terms. alpha = 1 is lasso, alpha = 0 ridge)</dd>


<dt>penaltyFunctionArguments</dt>
<dd>list with fields regularizedParameterLabels (labels of regularized parameters), and eps (controls the smooth approximation of non-differential penalty functions (e.g., lasso, adaptive lasso, or elastic net). Smaller values result in closer approximation, but may also cause larger issues in optimization.)</dd>

</dl>

.smoothElasticNetHessian: .smoothElasticNetHessian

Description

Usage

Value

Arguments