Ridge Estimator Bias

Problem

Return the bias of the ridge slope relative to OLS: Sxy/(Sxx + lambda) - Sxy/Sxx. Return 0.0 when Sxx or Sxx + lambda is zero.

Hint

Ridge always shrinks the slope toward zero, so the bias is negative for positive Sxy.

case 1

in: [10,5,5]

expected: -0.16666666666666669

case 2

in: [10,5,0]

expected: 0

case 3

in: [10,5,10]

expected: -0.25

case 4

in: [0,5,5]

expected: 0

case 5

in: [4,8,4]

expected: -1

Starter codePython

def ridge_bias(sxx, sxy, lam):
    # Your code here
    pass

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