Invariance Delta

Problem

Return the change in chi-square per degree of freedom between two nested measurement models: (chi2_metric - chi2_config)/(df_metric - df_config). Return 0.0 when the degrees-of-freedom difference is not positive.

Hint

A large chi-square change per df signals that the constraint hurts fit.

Test cases

case 1

in: [30,10,45,15]

expected: 3

case 2

in: [30,10,30,10]

expected: 0

case 3

in: [20,8,40,18]

expected: 2

case 4

in: [50,20,45,25]

expected: -1

case 5

in: [10,5,30,4]

expected: 0

Starter codePython

def invariance_delta(chi2_config, df_config, chi2_metric, df_metric):
    # Your code here
    pass

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