get_BxBy: Computes decomposition elements.

In LassoNet: 3CoSE Algorithm

Description

Computes matrices B^{ij}_X and B^{ij}_y to speed up estimation of connection signs. These matrices are stored only for indices that have non zero entries in penalty matrix M.

Usage

get.BxBy(x, y, M)

Arguments

x	Input data matrix of size n \times p, n - number of observations, p - number of covariates
y	y Response vector or size n \times 1
M	penalty matrix

Details

Calculates matrices all for i and j indices that have non zero values in a given penalty matrix.

Value

Bx	array of B^{ij}_X stored matrices. Bx[,,k] are the k-th combination of i and j non zero entry in the penalty matrix M
By	array of B^{ij}_y stored matrices. By[,k] are the k-th combination of i and j non zero entry in the penalty matrix M

Author(s)

Maintainer: Jonas Striaukas <[e-mail]>

References

Weber, M., Striaukas, J., Schumacher, M., Binder, H. "Network-Constrained Covariate Coefficient and Connection Sign Estimation" (2018) <doi:10.2139/ssrn.3211163>

Examples

p<-200
n<-100
x<-matrix(rnorm(n*p),n,p)
y<-rnorm(n,mean=0,sd=1)
M<-diag(p)
get.BxBy(x, y, M)

LassoNet documentation built on Jan. 19, 2020, 5:06 p.m.