To perform Adaptive lasso, which is the best package: msgps, or glmnet or glmaag? What's the difference among the three packages: algorithm, results?
Asked
Active
Viewed 257 times
0
-
1Welcome to CV! Have you tried using any of them yourself? Were there any discrepancies? One of the answers to [this related question](https://stats.stackexchange.com/q/231643/176202) shows a worked out example in `glmnet`. – Frans Rodenburg Jul 10 '19 at 02:58
-
Thanks for your advice. @FransRodenburg I've just tried them with the following case: n=20, p=152 (p>n). The results show that, msgps works well, while glmnet breaks down. As for glmaag, the one considering two tuning parameters, seems like a bit of overkill. – DreamZhang Jul 11 '19 at 02:26