Partial Dependence 1D

Problem

Compute the 1D partial dependence of a linear model: for each grid value v, average w[feature]*v plus the other weighted features over all rows.

Return [] if X or the grid is empty.

Hint

Partial dependence marginalizes the model over the other features.

case 1

in: [[[1,2],[2,3]],[1,1],0,[0,10]]

expected: [2.5,12.5]

case 2

in: [[[1,2],[2,3]],[1,1],1,[0,10]]

expected: [1.5,11.5]

case 3

in: [[],[1,1],0,[0,10]]

expected: []

case 4

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

expected: []

Starter codePython

def partial_dependence(X, w, feature, grid_values):
    # Your code here
    pass

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