Examples

These examples use the sparse group lasso and synthetic data to demonstrate how groupyr can be used to solve penalized regression problems with grouped covariates. Because the data is synthetically generated, we know the “ground truth” and can compare to the estimated important features chosen by the groupyr models.

Logistic Sparse Group Lasso for grouped sparse signals

Logistic Sparse Group Lasso for grouped sparse signals

Visualizing the SGL regularization path

Visualizing the SGL regularization path

Rudimentary performance comparison of groupyr and group-lasso

Rudimentary performance comparison of groupyr and group-lasso

Sparse Group Lasso for grouped sparse signals

Sparse Group Lasso for grouped sparse signals

Specifying a scoring criterion for model selection

Specifying a scoring criterion for model selection