filter_nonzero_bases

The <code>filter_nonzero_bases</code> function filters out bases functions that are essentially zero.
Bases functions with a sum of absolute values less than a threshold (\(10^{-10}\)) are considered as essentially zero and are filtered out.

Implements a partial linear semiparametric mixed-effects model (PLSMM) featuring a random intercept and applies a lasso penalty to both the fixed effects and the coefficients associated with the nonlinear function.
The model also accommodates interactions between the nonlinear function and a grouping variable, allowing for the capture of group-specific nonlinearities. Nonlinear functions are modeled using a set of bases functions. Estimation is conducted using a penalized Expectation-Maximization algorithm, and the package offers flexibility in choosing between various information criteria for model selection.
Post-selection inference is carried out using a debiasing method, while inference on the nonlinear functions employs a bootstrap approach.

Sami Leon

plsmmLasso

Variable Selection and Inference for Partial Semiparametric
Linear Mixed-Effects Model

Tong Tong Wu

filter_nonzero_bases function

<dl><dt>bases</dt>
<dd>A matrix containing the bases functions.</dd></dl>

Arguments

Filter bases functions — filter_nonzero_bases

<dl>

<dt>bases</dt>
<dd>A matrix containing the bases functions.</dd>

</dl>

filter_nonzero_bases: Filter bases functions

Description

Usage

Value

Arguments

Examples