DOCUMENTATION
Sign In

Get access to My SAS, trials, communities and more.

Machine Learning Procedures
 

GVARCLUS Procedure

Selection

To select the optimal lamda in the penalized log likelihood of the graphical lasso, PROC GVARCLUS minimizes the following fit statistics over all the steps,

minus log det ModifyingAbove normal upper Theta With caret plus normal upper T normal r normal a normal c normal e left-parenthesis upper S ModifyingAbove normal upper Theta With caret right-parenthesis plus sigma-summation Underscript 1 less-than-or-equal-to i less-than j less-than-or-equal-to p Endscripts upper I left-parenthesis ModifyingAbove theta With caret Subscript i j Baseline not-equals 0 right-parenthesis

where ModifyingAbove normal upper Theta With caret equals left-parenthesis ModifyingAbove theta With caret Subscript i j Baseline right-parenthesis Subscript 1 less-than-or-equal-to i comma j less-than-or-equal-to p denotes the estimates that are obtained at each step.

Last updated: September 04, 2026