glmnet_offset: Fit Penalized Generalized Linear Models with an Offset

In offsetreg: An Extension of 'Tidymodels' Supporting Offset Terms

Fit Penalized Generalized Linear Models with an Offset

Description

This function is a wrapper around glmnet::glmnet() that uses a column from x as an offset.

Usage

glmnet_offset(
  x,
  y,
  family,
  offset_col = "offset",
  weights = NULL,
  lambda = NULL,
  alpha = 1
)

Arguments

x	Input matrix
y	Response variable
family	A function or character string describing the link function and error distribution.
offset_col	Character string. The name of a column in data containing offsets.
weights	Optional weights to use in the fitting process.
lambda	A numeric vector of regularization penalty values
alpha	A number between zero and one denoting the proportion of L1 (lasso) versus L2 (ridge) regularization. alpha = 1: Pure lasso model alpha = 0: Pure ridge model

Details

Outside of the tidymodels ecosystem, glmnet_offset() has no advantages over glmnet::glmnet() since that function allows for offsets to be specified in its offset argument.

Within tidymodels, glmnet_offset() provides an advantage because it will ensure that offsets are included in the data whenever resamples are created.

The x, y, family, lambda, alpha and weights arguments have the same meanings as glmnet::glmnet(). See that function's documentation for full details.

Value

A glmnet object. See glmnet::glmnet() for full details.

See Also

glmnet::glmnet()

Examples

if (interactive()) {
  us_deaths$off <- log(us_deaths$population)
  x <- model.matrix(~ age_group + gender + off, us_deaths)[, -1]
  glmnet_offset(x, us_deaths$deaths, family = "poisson", offset_col = "off")
}

offsetreg documentation built on April 4, 2025, 1:35 a.m.