A group-wise descent method for fitting the standardized group lasso. The group lasso should be used when features have a natural grouping (eg. categorial variables) and group-wise sparsity is suspected (but within gruop sparsity is not). This package can also fit the ridged-group lasso, when additional within-group regularization is needed (eg. if group sizes exceed the number of observations). Can be used for linear and logistic models.