find_independent_regression_vectors.glmnet_multivariate: Given a regularization path, find K selected independant variables.
Description
find_independent_regression_vectors tries to find a situation where K variables
are selected, so that the selected columns of matrix Xp are independant.
integer, position of the root column (intercept) excluded from the fit. null
if no root column.
Value
delta a vector of regression with K non-zero coefficients.
Details
To do that, if a set of selected is not independent, we go back to the previous selected
variables, and forbid the moves that led to non-independance for the rest of the path.