GIN Sum vs Mean Check

Problem

Check whether sum pooling and mean pooling can distinguish two bags of node features.

features_a and features_b are lists of numbers representing each graph's node features. Return [s, m] where s is 1 when the sums differ and m is 1 when the means differ, else 0.

Hint

A graph neural network with mean readout cannot tell apart equal-size bags with equal means.

Test cases

case 1

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

expected: [0,1]

case 2

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

expected: [0,0]

case 3

in: [[1],[2]]

expected: [1,1]

case 4

in: [[],[]]

expected: [0,0]

Starter codePython

def gin_sum_vs_mean_check(features_a, features_b):
    # Your code here
    pass

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