Closed ppttzhu closed 5 years ago
aronwc
commented
5 years ago The instructions in partition_girvan_newman:
If there are ties in edge betweenness, break by edge name (e.g., (('A', 'B'), 1.0) comes before (('B', 'C'), 1.0)).
Yes, sorting by the edge names is sufficient. E.g.:
>>> sorted([('A', 'C'), ('B', 'A'), ('A', 'B')])
[('A', 'B'), ('A', 'C'), ('B', 'A')]Note that when you sort tuples, it first sorts by the first element of each tuple, then the second, etc.
ppttzhu
commented
5 years ago Great, that's helpful.
The instructions in partition_girvan_newman:
If there are ties in edge betweenness, break by edge name (e.g., (('A', 'B'), 1.0) comes before (('B', 'C'), 1.0)).
Yes, sorting by the edge names is sufficient. E.g.:
>>> sorted([('A', 'C'), ('B', 'A'), ('A', 'B')]) [('A', 'B'), ('A', 'C'), ('B', 'A')]Note that when you sort tuples, it first sorts by the first element of each tuple, then the second, etc.
sorted graph.pdf
In norm_cut_score test case max_depth=1, I cannot get 1.007. I think the difference is caused by tie breaking rule. HW instruction is ambiguous:
break by edge name (e.g., (('A', 'B'), 1.0) comes before (('B', 'C'), 1.0)).It should be more clear such as: alphabetical order by first_node then second_node, given an edge (first_node, second_node).If I use python3 default sort() function, the sequence I get is as attached.