AIC from Residuals

Problem

Return the Akaike information criterion computed from the residual sum of squares for a Gaussian model:

AIC = n * ln(RSS / n) + 2k

using the natural logarithm, with n samples and k parameters (including the variance).

Hint

The penalty 2k grows linearly with the number of parameters.

case 1

in: [10,100,2]

expected: -226.25850929940455

case 2

in: [100,100,2]

expected: 4

case 3

in: [1,10,1]

expected: -21.025850929940454

case 4

in: [400,100,5]

expected: 148.62943611198907

Starter codePython

import math
def aic_from_rss(rss, n, k):
    # Your code here
    pass

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