Condition Number

Problem

Compute the condition number of a design matrix X with two columns: sqrt(lambda_max / lambda_min) where the lambdas are the eigenvalues of X^T X.

Return 0.0 when the smaller eigenvalue is non-positive.

Hint

Large condition numbers mean the normal equations are poorly conditioned.

case 1

in: [[[1,0],[0,1]]]

expected: 1

case 2

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

expected: 2

case 3

in: [[[1,1],[1,1]]]

expected: 0

case 4

in: [[[1,1],[1,2]]]

expected: 6.854101966249688

Starter codePython

def condition_number(X):
    # Your code here
    pass

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