MediumCalculusca-211

Proximal Operator of Elastic Net

Problem

The elastic net penalty is lambda1 * ||v||_1 + (lambda2 / 2) * ||v||_2^2, combining sparsity with shrinkage.

Its proximal operator applies soft-thresholding and then divides by (1 + lambda2): prox(v_i) = soft(v_i, lambda1) / (1 + lambda2). Given the values and both penalties, return the result.

Hint

The L2 part rescales all surviving coefficients uniformly.

Test cases

case 1

in: [[3,-2],0.5,1]

expected: [1.25,-0.75]

case 2

in: [[1],1,0]

expected: [0]

case 3

in: [[4],0,2]

expected: [1.3333333333333333]

case 4

in: [[0.3],0.5,1]

expected: [0]

Starter codePython

def prox_elastic_net(values, lam1, lam2):
    # Your code here
    pass

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