How are the bounds for the regularization parameter lambda computed?

[Timvanderschaaf]

I cannot figure out what the reasoning is behind the bounds for the regularization parameter lambda in the Ridge Regression.

Please see below snippets of code from model.R. The first code snippet (see also last snippet) uses a function to compute a list of lambdas, however...

... in the model_refit function, uses lambda_hp and I cannot find out where this value comes from?

I would like to understand how the list of possible lambda values exactly is being constructed.

[gufengzhou]

Sorry for the late reply. lambda_seq is derived directly from the glmnet package. Not sure about your question tho.