grpCox: Penalized Cox Model for High-Dimensional Data with Grouped Predictors

Fit the penalized Cox models with both non-overlapping and overlapping grouped penalties including the group lasso, group smoothly clipped absolute deviation, and group minimax concave penalty. The algorithms combine the MM approach and group-wise descent with some computational tricks including the screening, active set, and warm-start. Different tuning regularization parameter methods are provided.

install.packages("grpCox")

Try the grpCox package in your browser