Elastic Net Objective

Problem

Compute the elastic net objective for linear regression without an intercept: (1/(2n)) * ||Xw - y||^2 + alpha * l1_ratio * sum(|w_j|) + alpha * (1 - l1_ratio) / 2 * sum(w_j^2).

Empty X returns 0.0.

Hint

l1_ratio blends the L1 and L2 penalties; the first term is the mean squared error over 2.

Test cases

case 1

in: [[[1],[2]],[2,4],[2],0.1,0.5]

expected: 0.2

case 2

in: [[[1],[2]],[1,2],[0],0.1,0.5]

expected: 1.25

case 3

in: [[[1,0],[0,1]],[1,1],[1,-1],0.2,0.5]

expected: 1.3

case 4

in: [[],[],[1],0.1,0.5]

expected: 0

Starter codePython

def elastic_net_objective(X, y, w, alpha, l1_ratio):
    # Your code here
    pass

4 test cases · run your code against them in the editor