search

archer-yang-lab / gglasso

Automatically exported from code.google.com/p/gglasso
GNU General Public License v2.0
10 stars 5 forks source link

Class of y argument and logistic regression #12

Open ArthurAndrews opened 4 years ago

ArthurAndrews commented 4 years ago

Respectfully, I think there may be an error in the gglasso and cv.gglasso documentation. The documentation specifies that y should be a two-level factor for classification. However, gglasso throws an error if y is any class other than numeric, and the example given for logistic regression with the colon data set supplies a numeric y with values of -1 and 1.

Thanks for the great package!

jakobdambon commented 2 years ago

Yes, I've encountered the same issue today.