pengls

pengls fits penalised generalised least squares models that handle both autocorrelation in high dimensional datasets and high dimensionality through an iterative combination of generalised least squares and penalised least squares methods.

Key Features:

• Autocorrelation handling: Incorporates generalised least squares methodology from the nlme package to model autocorrelation structures in data.
• High dimensionality handling: Integrates penalised least squares methods from the glmnet package for variable selection and regularisation in high dimensional settings.
• Iterative algorithm: Combines both methodologies through an iterative loop that alternates between estimating autocorrelation parameters and penalised regression coefficients.

Scientific Applications:

• Spatial data analysis: Applicable to high dimensional datasets that exhibit spatial autocorrelation patterns.
• Temporal data analysis: Applicable to high dimensional datasets that exhibit temporal autocorrelation patterns.

Methodology:

The method combines generalised least squares from the nlme package with penalised least squares from the glmnet package through an iterative loop.