Lasso or ridge regularization for least squares in LAPACK?

[Kadence]

Can one perform regularization for least squares using xgels or another routine in LAPACK, i.e. placing constraints or penalties on the solution (x) vectors in Ax=B?

[Kadence]

Trying once more to see if this is possible.

[sven]

I don't think that it is straightforward to do Lasso regression using LAPACK. If I understand correctly, it is a constrained least squares problem. You could use xGELS to help with simple damped regression, but you would have to use some method yourself to choose the damping parameter, lambda, in (A - lambda*I).

Sorry not to be of more help,

Sven Hammarling.