listToMatrix: Obtain a sparse matrix of the coefficients of the path

In MLGL: Multi-Layer Group-Lasso

Obtain a sparse matrix of the coefficients of the path

Description

Obtain a sparse matrix of the coefficients of the path

Usage

listToMatrix(x, row = c("covariates", "lambda"))

Arguments

x	MLGL object
row	"lambda" or "covariates". If row="covariates", each row of the output matrix represents a covariate else if row="lambda", it represents a value of lambda.

Details

This function can be used with a MLGL object to obtain a matrix with all estimated coefficients for the p original variables. In case of overlapping groups, coefficients from repeated variables are summed.

Value

a sparse matrix containing the estimated coefficients for different lambdas

See Also

MLGL, overlapgglasso

Examples

# Simulate gaussian data with block-diagonal variance matrix containing 12 blocks of size 5
X <- simuBlockGaussian(50, 12, 5, 0.7)
# Generate a response variable
y <- X[, c(2, 7, 12)] %*% c(2, 2, -2) + rnorm(50, 0, 0.5)
# Apply MLGL method
res <- MLGL(X, y)
# Convert output in sparse matrix format
beta <- listToMatrix(res)

MLGL documentation built on March 31, 2023, 9:32 p.m.