Population Attributable Fraction

Problem

Return the population attributable fraction p_e*(RR - 1)/(1 + p_e*(RR - 1)), where p_e = (a+b)/n is the exposure prevalence and RR is the risk ratio. Return 0.0 when the table is empty or the unexposed risk is zero.

Hint

PAF combines the exposure prevalence with the strength of the risk ratio.

Test cases

case 1

in: [30,70,10,90]

expected: 0.49999999999999994

case 2

in: [20,80,20,80]

expected: 0

case 3

in: [50,50,10,90]

expected: 0.6666666666666666

case 4

in: [10,90,20,80]

expected: -0.3333333333333333

case 5

in: [0,10,5,5]

expected: -1

Starter codePython

def population_attributable_fraction(a, b, c, d):
    # Your code here
    pass

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