MediumTime Seriests-166

Nash-Sutcliffe Efficiency

Problem

Compute the Nash-Sutcliffe efficiency: 1 - SSE(model) / SSE(mean), where SSE(mean) uses the mean of actual.

Return 0.0 when actual is empty or has zero variance.

Hint

Using the mean of actual as the benchmark is equivalent to R squared.

case 1

in: [[1,2,3,4],[1,2,3,4]]

expected: 1

case 2

in: [[1,2,3,4],[2,2,2,2]]

expected: -0.19999999999999996

case 3

in: [[3,3,3],[3,3,3]]

expected: 0

case 4

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

expected: -3

Starter codePython

def nash_sutcliffe(actual, forecast):
    # Your code here
    pass

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