Open biona001 opened 2 years ago
It looks like the method that supports the string-vector input is this one:
So this works:
using GLMNet
y = ["M", "B", "M", "B"]
X = rand(4, 10)
glmnet(X, y)
The reason it doesn't need a distribution is because it chooses between Binomial and Multinomial based on the number of unique values in y
. This method could probably be extended to support passing a distribution, and I guess throwing an error if the distribution and y
are incompatible.
At the very least the README should be updated
From README:
But the following doesn't work
Fortunately if
y
is a matrix with 2 columns, it does work