grpregOverlap (version 2.2-0)

cv.grpregOverlap: Cross-validation for choosing regularization parameter lambda

Description

Performs k-fold cross validation for penalized regression models with overlapping grouped covariates over a grid of values for the regularization parameter lambda.

Usage

cv.grpregOverlap(X, y, group, ..., nfolds = 10, seed, trace = FALSE)

Arguments

X

The design matrix, without an intercept, as in grpregOverlap.

y

The response vector (or matrix), as in grpregOverlap.

group

A list of vectors containing group information, as in grpregOverlap.

...

Additional arguments to grpregOverlap.

nfolds

The number of cross-validation folds. Default is 10.

seed

Set the seed of the random number generator to obtain reproducible results.

trace

If set to TRUE, print out the progress of the cross-validation. Default is FALSE.

Value

An object with S3 class "cv.grpregOverlap", which inherits from class "cv.grpreg". The following variables are contained in the class (adopted from cv.grpreg).

Details

This function is built upon cv.grpreg. The class can directly call plot function implemented for class cv.grpreg.

References

• Breheny P (2014). R package 'grpreg'. https://CRAN.R-project.org/package=grpreg/grpreg.pdf

See Also

grpregOverlap, predict.grpregOverlap, summary, and cv.grpreg.

Examples

## linear regression, a simulation demo.
set.seed(123)
group <- list(gr1 = c(1, 2, 3),
              gr2 = c(1, 4),
              gr3 = c(2, 4, 5),
              gr4 = c(3, 5),
              gr5 = c(6))
beta.latent.T <- c(5, 5, 5, 0, 0, 0, 0, 0, 5, 5, 0) # true latent coefficients.
# beta.T <- c(5, 5, 10, 0, 5, 0), true variables: 1, 2, 3, 5; true groups: 1, 4.
X <- matrix(rnorm(n = 6*100), ncol = 6)
X.latent <- expandX(X, group)
y <- X.latent %*% beta.latent.T + rnorm(100)

cvfit <- cv.grpregOverlap(X, y, group, penalty = 'grMCP')
summary(cvfit)
plot(cvfit)
par(mfrow=c(2,2))
plot(cvfit, type="all")