fit.LASSO: Penalized backend fitting wrappers

In peperr: Parallelised Estimation of Prediction Error

Penalized backend fitting wrappers

Description

Interface functions for fitting penalized models in peperr. The generic wrapper is fit.penalized; fit.LASSO and fit.fusedLASSO are convenience aliases for plain and fused lasso fits.

Usage

fit.penalized(response, x, cplx, ...)
fit.LASSO(response, x, cplx, ...)
fit.fusedLASSO(response, x, cplx, ...)

Arguments

response	response. This can be a binary vector, a numeric response, a Surv object, or a two-column matrix with time and status.
x	n * p matrix of covariates.
cplx	complexity value. For the penalized backend this is usually lambda1, or a named list of tuning arguments.
...	additional arguments passed to penalized.

Details

fit.penalized is the general wrapper around penalized. For survival models it also stores the training design matrix and response so that predictProb.penfit and PLL.penfit can be used later by pmpec and peperr.

fit.LASSO simply calls fit.penalized. fit.fusedLASSO calls the same backend with fusedl = TRUE.

Value

A fitted penfit object.

See Also

complexity.cv.penalized, predictProb.penfit, PLL.penfit, penalized

peperr documentation built on March 25, 2026, 9:06 a.m.