ps16c.py

from os import listdir
from os.path import isfile, join
import multiprocessing
import networkx as nx
from sklearn.utils.graph import single_source_shortest_path_length as sssp


fb_folder = "just8"

def func_sssp(component):
    am = nx.adjacency_matrix(component)
    length = len(component.nodes())
    geodesic_vals = []
    for idx in range(length):
        sssp_dicts = sssp(am, int(idx))
        filtered = filter(lambda a: a != 0, sssp_dicts.values())
        geodesic_vals.extend(filtered)
        #geodesic_vals.extend(sssp_dicts.values())
    return geodesic_vals

def get_max_component(g):
    components = nx.connected_component_subgraphs(g)
    return components.next()

def do(fb_file):
    g = nx.read_edgelist(join(fb_folder,fb_file))
    giant_component = get_max_component(g)
    geodesic_vals = func_sssp(giant_component)
    return geodesic_vals

fb_files = [ f for f in listdir(fb_folder) if isfile(join(fb_folder,f)) ]
p = multiprocessing.Pool(multiprocessing.cpu_count())
output_list = p.map(do, fb_files)
grand_total = 0
total_num = 0
max_val = 0
for output in output_list:
    grand_total += sum(output)
    total_num += len(output)
    max_output = max(output)
    max_val = max(max_val, max_output)

mean = grand_total / total_num
print "MEAN:", mean
print "MAX :", max_val

I will run this code on my machine asap and I will share the result. Thanks!