The pyg implement of SSGC

[Haldate-Yu]

Hi, I noticed an inconsistency between the paper and this implementation. Equation 13 in the paper, x_final = alpha * x_ori + (1 - alpha) * mean(x_list); But in the line 101 of ssgc-mlp.py, x_final = alpha * x_ori + mean(x_list). Could you tell me what causes the difference?

[allenhaozhu]

I guess the difference is not a issue for the final result. Because it is more like a global scale (very close to 1) to the final features. My fault.