RGCCA with L1 sparsity for variable selection.
Usage
sgcca(
A,
C = 1 - diag(length(A)),
c1 = rep(1, length(A)),
ncomp = rep(1, length(A)),
scheme = "centroid",
scale = TRUE,
init = "svd",
bias = TRUE,
tol = .Machine$double.eps,
verbose = FALSE
)
Arguments
- A
List of data blocks.
- C
Design matrix.
- c1
L1 penalty (between 0 and 1).
- ncomp
Number of components per block.
- scheme
Scheme type.
- scale
Scale blocks.
- init
Initialization method.
- bias
Biased estimator.
- tol
Tolerance.
- verbose
Print progress.
Value
List with Y, a, astar, C, c1, scheme, ncomp, crit, AVE.