Importing and visualizing Twitter networks

Analyzing Social Media Data in Python

Alex Hanna

Computational Social Scientist

Edge Lists

         BethMohn           ChristianMohn
           ASilNY         LarrySchweikart
         mattg444              WhiteHouse
    hlthiskrieger                aravosis
          Herky86          SenJeffMerkley
  PatrickParsons9              TwitterGov
    New_Narrative                 CFR_org
           dddlor               roywoodjr
      scrivener50          michaelscherer
  ChiefsHeadCoach           johnpavlovitz

Importing a retweet network

import networkx as nx

## ... flatten and convert JSON

G_rt = nx.from_pandas_edgelist(
                        tweets,
                        source = 'user-screen_name', 
                        target = 'retweeted_status-user-screen_name',
                        create_using = nx.DiGraph())

Importing a quoted network

import networkx as nx

## ... flatten and convert JSON

G_quote = nx.from_pandas_edgelist(
                           tweets,
                           source = 'user-screen_name', 
                           target = 'quoted_status-user-screen_name',
                           create_using = nx.DiGraph())

Importing a reply network

import networkx as nx

## ... flatten and convert JSON

G_reply = nx.from_pandas_edgelist(
                           tweets,
                           source = 'user-screen_name', 
                           target = 'in_reply_to_screen_name'
                           create_using = nx.DiGraph())

Visualization

nx.draw_networkx(T)
plt.axis('off')

Directed graph

Visualization options

sizes = 
   [x[1]*100 for x in T.degree()]
nx.draw_networkx(T, 
            node_size = sizes, 
            with_labels = False, 
            alpha = 0.6,
            width = 0.3)
plt.axis('off')

Directed graph modified

Circular layout

circle_pos =
    nx.circular_layout(T)
nx.draw_networkx(T,
            pos = circle_pos,
            node_size = sizes, 
            with_labels = False, 
            alpha = 0.6,
            width = 0.3)
plt.axis('off')

Directed graph circular layout

Let's practice!

Analyzing Social Media Data in Python