Introduction to the dataset

Intermediate Network Analysis in Python

Eric Ma

Data Carpentry instructor and author of nxviz package

Dataset & case study introduction

College forum posting dataset, 6 months
Node partitions: students, forums
Activities in the chapter:
- Constructing a graph from a pandas DataFrame
- Computing unipartite projections of a bipartite graph
- Visualization
- Time series filtering & analysis
Recap previously used functions

Graphs from DataFrames

df

   customers    products
0  customerA    product1
1  customerB    product2
...

list(G = nx.Graph())

G.add_nodes_from(df['products'], bipartite='products')
G.add_nodes_from(df['customers'], bipartite='customers')

list(G.nodes())

['product1', 'customerC', 'product2', 'customerB', 'customerA']

list(G.edges())

[]

Graphs from DataFrames

G.add_edges_from(zip(df['customers'], df['products']))

list(G.edges())

[('product1', 'customerC'), ('product1', 'customerA'), 
    ('customerC', 'product2'), ('product2', 'customerB')]

Bipartite projections

cust_nodes = [n for n in G.nodes() if G.node[n]
                 ['bipartite'] == 'customers']

prod_nodes = [n for n in G.nodes() if G.node[n]
                  ['bipartite'] == 'products']

prodG = nx.bipartite.projected_graph(G, nodes=prod_nodes)
custG = nx.bipartite.projected_graph(G, nodes=cust_nodes)

list(prodG.nodes())

['product1', 'product2']

list(custG.nodes())

['customerC', 'customerB', 'customerA']

Let's practice!

Intermediate Network Analysis in Python