Standard error for ridge regression coefficients computed incorrectly

[dlaprins]

In https://github.com/ing-bank/skorecard/blob/main/skorecard/linear_model/linear_model.py, the standard errors for the coefficients of logistic regression are computed (lines 132-142). For ridge logistic regression, I believe this formula is incorrect: see for example page 12 of [https://arxiv.org/pdf/1509.09169.pdf](these lecture notes).

Similarly I believe the expression is incorrect for lasso logistic regression; I am unaware of an exact expression for the standard error in this case (and by extension, elasticnet).