dot-adjustX function

Adjust a model matrix When not rank deficient, do nothing. When rank deficient matrix, drop columns.

.adjustX(X, tol = NULL, why_dropped = FALSE)

Arguments

• X: model matrix
• tol: non-negative tolerance for testing for "practically zero" singular values (passed to Matrix::rankMatrix())
• why_dropped: logical indicating whether or not to provide information about sets of collinear predictors (not yet implemented)