Extract one model from a fitted "abess" object. — extract • abess

Extract the fixed-support-size model's information such as the selected predictors, coefficient estimation, and so on.

extract(object, support.size = NULL, ...)

# S3 method for class 'abess'
extract(object, support.size = NULL, ...)

Arguments

object: An "abess" project.
support.size: An integer value specifies the model size fitted at given support.size. If support.size = NULL, then the model with the best tuning value would be returned. Default: support.size = NULL.
...: Other arguments.

Value

A list object including the following components:

beta: A \(p\)-by-1 matrix of sparse matrix, stored in column format.
intercept: The fitted intercept value.
support.size: The support.size used in the function.
support.beta: The support.size-length vector of fitted coefficients on the support set.
support.vars: The character vector gives variables in the support set.
tune.value: The tuning value of the model.
dev: The deviance of the model.

See also

print.abess, predict.abess, coef.abess, extract.abess, plot.abess, deviance.abess.