def test_nx_cluster(self):
g = nx.karate_club_graph()
coms = algorithms.louvain(g)
pos = nx.spring_layout(g)
viz.plot_network_clusters(g, coms, pos)
plt.savefig("cluster.pdf")
os.remove("cluster.pdf")
coms = algorithms.angel(g, 0.25)
pos = nx.spring_layout(g)
viz.plot_network_clusters(g,
coms,
pos,
plot_labels=True,
plot_overlaps=True)
plt.savefig("cluster.pdf")
os.remove("cluster.pdf")
#creates graph G = nx.Graph(filtered_entities) print(G.nodes()) print(G.edges()) #remove outliers/self-loops G.remove_edges_from(nx.selfloop_edges(G)) G = nx.k_core(G,k=2) #Louvain/infomap algorithm and graph plot #coms = algorithms.louvain(G) coms = algorithms.infomap(G) pos = nx.spring_layout(G) viz.plot_community_graph(G, coms, figsize=(8, 8), node_size=200, plot_overlaps=False, plot_labels=True, cmap=None, top_k=None, min_size=None) viz.plot_network_clusters(G, coms, position=None, figsize=(8, 8), node_size=200, plot_overlaps=False, plot_labels=False, cmap=None, top_k=None, min_size=None) #converting this to an nx graph for calculations. mod = evaluation.modularity_density(G,coms) print(mod) #calculating modularity mod = evaluation.modularity_density(G,coms) print(mod) #calculating purity #communities = eva(G, coms) #pur = evaluation.purity(communities) #print(pur) #calculating avg embeddedness
plt.show()
#%%
evaluation.newman_girvan_modularity(unet,algorithms.infomap(net))
#%%
import networkx as nx
from cdlib import algorithms, viz
coms = algorithms.infomap(net)
pos = nx.spring_layout(net)
viz.plot_network_clusters(net, coms, pos,plot_labels=True)
#%%
#not in book
from heapq import nlargest
h,a=nx.hits(net)
hubs=nlargest(2, h, key = h.get)
labels = {}
for node in net.nodes():
if node in hubs:
#set the node name as the key and the label as its value
labels[node] = node
# find all the unique communities and define an empty set to catch a frozenset of all nodes in each community
truth_coms = df_lcc_truth['truth'].unique()
communities_set = set()
# iterate through the unique communities to get the nodes in each community and add them to the set as frozensets
for com in truth_coms:
com_nodes = df_lcc_truth[df_lcc_truth['truth'] == com].iloc[:, 0].values
communities_set.add(frozenset(com_nodes))
# build the nodeclustering object
ground_truth_com = NodeClustering(communities=communities_set, graph=nx_g, method_name="ground_truth")
#%% define and plot ground truth
# define the positions here so all the cluster plots in the loop are the same structure
pos = nx.fruchterman_reingold_layout(nx_g)
#%%
# plot the original network with the ground truth communities
viz.plot_network_clusters(nx_g, ground_truth_com, pos, figsize=(5, 5))
#nx.draw_networkx_labels(nx_g, pos=pos)
plt.title(f'Ground Truth of {graph_name}')
plt.show()
#%% evaluate ground communitiy metrics
viz.plot_com_properties_relation(ground_truth_com, evaluation.size,
evaluation.internal_edge_density)
plt.show()
#%%
viz.plot_com_stat(ground_truth_com, evaluation.conductance)
plt.show()
viz.plot_com_stat(ground_truth_com, evaluation.average_internal_degree)
plt.show()
for c in range(len(communities)):
for i in communities[c]:
coms_dict[i] = [c]
# make a df with the results of the algo
df_clusts = pd.DataFrame.from_dict(coms_dict).T.reset_index()
df_clusts.columns = ['node', name]
# merge this results with the df_nodes to keep track of all the nodes' clusters
df_nodes = pd.merge(df_nodes,
df_clusts,
how='left',
left_on='node',
right_on='node')
# plot the network clusters
viz.plot_network_clusters(nx_g, pred_coms, pos, figsize=(5, 5))
plt.title(f'{name} algo of {graph_name}')
plt.show()
# plot the graph
viz.plot_community_graph(nx_g, pred_coms, figsize=(5, 5))
plt.title(f'Communities for {name} algo of {graph_name}.')
plt.show()
#%%
odd_ports = ['ATLANTIC CITY', 'OCEAN CITY', 'KEY WEST']
df_odd_ports = df_edgelist[(df_edgelist['Source'].isin(odd_ports)) |
(df_edgelist['Target'].isin(odd_ports))]
#%% explore communities in communities
if i == j:
count_leiden += 1
l.remove(i)
print("Equal community:", count_leiden)
print("#total communtites in leiden:", len(coms_leiden.communities))
print("#total communtites in graph:", len(coms_graph))
# Visualising a subset of nodes
# Subset
import numpy as np
F = nx.Graph()
for (u, v, wt) in G.edges.data('weight'):
if wt > 500:
print(u, v, wt)
if u not in F.nodes():
F.add_node(u,
labels=G.nodes()[u]['labels'],
properties=G.nodes()[u]['properties'])
if v not in F.nodes():
F.add_node(v,
labels=G.nodes()[v]['labels'],
properties=G.nodes()[v]['properties'])
F.add_edge(u, v, weight=wt)
pos = nx.spring_layout(F, k=10 / np.sqrt(len(F.nodes())), iterations=20)
viz.plot_network_clusters(F, coms, pos, plot_labels=True, node_size=400)
#if __name__ == '__main__':
#create_networkx_graph()`
def draw_community_graph(g, coms, filename):
pos = nx.spring_layout(g)
viz.plot_network_clusters(g, coms, pos)
plt.savefig(filename)
cs.erdos_renyi_modularity().score))
print("{0:>15s} | {1:.6f}".format(
'Robustness',
cs.normalized_mutual_information(alg(G)).score))
print("{0:>15s} | {1:.1f} sec\n".format('Timing', time() - tic))
return cs
# G = read('toy')
G = read('karate')
# G = read('women')
# G = read('dolphins')
# G = read('got-appearance')
# G = read('diseasome')
# G = read('wars')
# G = read('transport')
# G = read('java')
# G = read('imdb')
# G = read('wikileaks')
info(G)
clusters(G, lambda G: algorithms.girvan_newman(G, level=1))
clusters(G, lambda G: algorithms.label_propagation(G))
cs = clusters(G, lambda G: algorithms.louvain(G))
# clusters(G, lambda G: algorithms.leiden(G))
# clusters(G, lambda G: algorithms.sbm_dl(G))
viz.plot_network_clusters(G, cs, nx.spring_layout(G))
plt.show()