Enet: Elastic Net Model Fitting

In IOBR/IOBR: Immune Oncology Biological Research

Elastic Net Model Fitting

Description

Fits an elastic net model using the 'caret' package, allowing for both L1 (Lasso) and L2 (Ridge) regularization. The function uses cross-validation to identify the optimal alpha and lambda values that maximize accuracy. Alpha ranges from 0 (Ridge) to 1 (Lasso), with lambda controlling the strength of the regularization.

Usage

Enet(train.x, train.y, lambdamax, nfold = nfold)

Arguments

train.x	A matrix or data frame containing training predictors.
train.y	A numeric vector or factor representing the binary outcome for each sample in the training set.
lambdamax	The maximum value of lambda to consider in the regularization path.
nfold	The number of folds to use for cross-validation, which also determines the number of repeats for repeated CV.

Value

A list containing the optimal values of alpha and lambda chosen based on cross-validation.

Examples

# Assuming 'train_data' and 'train_outcome' are already defined:
train_data <- matrix(rnorm(100 * 10), ncol = 10)
train_outcome <- rbinom(100, 1, 0.5)
optimal_parameters <- Enet(train.x = train_data, train.y = train_outcome,
                           lambdamax = 1, nfold = 10)
print(optimal_parameters)

IOBR/IOBR documentation built on April 3, 2025, 2:19 p.m.