Metrics

# Import NetworkX and key data science libraries
import networkx as nx
import pandas as pd
import numpy as np
import altair as alt

# Import Marvel network data
marvel= pd.read_csv("../data/marvel-unimodal-edges.csv")
M = nx.from_pandas_edgelist(marvel, source="Source", target="Target", edge_attr=True)
print(M)

M.number_of_nodes() # Number of nodes

M.number_of_edges() # Alternative method for number of edges

M.is_directed() # Is the graph directed?

nx.shortest_path(M, "Black Panther / T'chal", 	"Gambit / Remy Lebeau") # Shortest path as list

nx.shortest_path_length(M, "Black Panther / T'chal", "Gambit / Remy Lebeau") #Shortest path length (Distance)

nx.average_shortest_path_length(M) # Average shortest path length (Average distance)

nx.diameter(M)

nx.is_connected(M) # Is the graph connected?

nx.number_connected_components(M) # How many components are there?

# Find the largest connected component (as a node set)
components = [G for G in nx.connected_components(M)]
largest = max(components, key=len)

# Calculate betweenness centrality for every node
bc = nx.betweenness_centrality(M)

# Get number of triangles for every node
nx.triangles(M)

# n.b. Pay attention to the following note from the NetworkX docs:
# When computing triangles for the entire graph each triangle is counted three times, once at each node.

# Get clustering coefficients
# These can be added to a node dataframe as in the previous chapter
nx.clustering(M)

# Find the average clustering coefficient
nx.average_clustering(M)

# List a network's bridges
list(nx.bridges(M))

# Test if a network has any bridges
nx.has_bridges(M)

# List a network's local bridges
list(nx.local_bridges(M))

# Find a node's neighbors
list(nx.neighbors(M, 'Death'))

# Find common neighbors of a node pair
list(nx.common_neighbors(M, 'Death', 'Thanos'))

# Get a list of the basis of all cycle in the graph
# All larger cycles are based on these cycles, many of which are triangles
nx.cycle_basis(M)