Tweet2Graph

Fast and easy social graphs from tweets

Installation

To install using pip, use

pip install tweet2graph

Usage

• Import the library:

from tweet2graph import Tweets2Graph

• Choose which interaction between users are relevant for you:

backend = Tweets2Graph(interactions=["retweet","quote","reply"], #relevant interactions
                          username="screen_name")                #name of the nodes

• Load dataset

#load a single .csv/.json file
backend.from_file("examples/csv/1614874276.csv")

#load a folder 
backend.from_folder("examples/csv/")

#load from a mongodb collections
backend.from_mongo(connection_string='<CONNECTION_STRING>',
                    database='<DB_NAME>',
                    collection='<COLLECTION_NAME>')
                    
#or from a pandas dataframe
backend.from_dataframe(dataframe)

• Fit/Transform

  #organize data
  backend.fit()
  
  #create graph
  graph = backend.transform()
  
  #or simply
  graph = backend.fit_transform()

• Show and save

  import networkx as nx
  nx.draw(graph,with_labels=False,node_size=5,font_size=5,font_color="red",node_color="black")
  nx.write_adjlist(G=graph)

TODO:

• add connected component
• .from_user('<USER_NAME>') method
• Add metadata to tweet
• Insert mentions in possible interactions
• Add custom error and better explanation
• Add speed benchmark and figures
• Add .from_stream(id="<hashtag>") method
• Update .from_stream(id="<hashtag>") with two distinct processes
• Dockerfile

License

MIT