grpss (version 3.0.1)

summary.cv.grpreg: Summarize a fitted cv.grpreg object

Description

Summarizes a fitted penalized regression model with 'cv.grpreg' class.

Usage

"summary"(object, digits = 4, ...)

Arguments

object

A fitted "cv.grpreg" object.

digits

Number of digits past the decimal point to print out. The default is 4.

...

Optional arguments passed to other methods.

Value

A list with class "summary.cv.grpreg" containing the following components:

n

Number of observations.

p

Number of screened predictors.

penalty

The penalty applied to the model.

model

The type of model.

family

The link function.

criterion

The screening criterion.

lambda

The default or specified regularization parameter.

beta

The estimates of coefficients at the specified lambda.

iter

The number of iterations at the specified lambda.

df

The estimates of effective number of model parameters at the specified lambda.

cve

The average cross-validation error at specified lambda.

call

The function call.

Additional elements are contained for the case in which family = "gaussian":

r.squared

The r.squared.

snr

The signal-to-noise ratio.

scale

The scale parameter estimate (sigma).

and following elements for the case in which family = "poisson" or "binomial":

logLik

The negative log-likelihood values for the fitted model.

aic

Akaike's information criterion (AIC).

bic

Bayesian information criterion (BIC).

aicc

The AIC with a correction for finite sample sizes (AICC).

pe

The prediction error for family = "binomial".

Details

This function is similar to summary.cv.grpreg in grpreg package, but gives several different results. The lambda here is only valid for "grpreg" object. See summary.grpreg.

See Also

grpss, summary.grpreg