hdvvip / cs224w_winter2021 Goto Github PK

I think the equation in the second image is the same as the equation in the first image.Therefore,you do not need to add (1 - B) / N in each loop.
Maybe you can alter your code just like this

for node_neighbor in G.neighbors(node_id):
    node_deg = G.degree[node_neighbor]
    # Notice that all nodes share the same PageRank value r0 at the first iteration
    r1 += beta * r0/node_deg
r1 += (1-beta)*1/N
    
r1 = round(r1, 2)

Do you think it would be possible to post the homework question pdf files? Thank you!

1. For Q4, networkx closeness centrality returns the normalized closeness directly, which is different from the raw (unnormalized) one that we learned in the lecture. You should add closeness = closeness * (G.number_of_nodes() - 1) in your code.
2. For Q6, you directly used edge_1 in is_neg_edge(), which causes your outputs to be wrong.

Thanks for sharing! I am interested in this course too, and I found that there are 3 homework assignments in this course. However these assignments seem to have no links on the course website. Is there any way to get them?

When I was running the following codes,

from deepsnap.hetero_graph import HeteroGraph

hete = HeteroGraph(G_orig)

I got the error, saying
TypeError: Unknown type color in edge attributes.

However, in the previous lines, there is a line like:
colors = nx.get_edge_attributes(G,'color').values()
and it works. It means G has the attributes 'color'. So I am confused, what should I do?

hello there i am just wondering why do we need a nega-sampling here, isn't nega-sampling only for random walk?