Python Social Media Analytics

Building a graph

In this section, we will follow a similar logic to create our graph, but we will engineer the nodes and edges differently to obtain the relationship that we are looking for.

We create a list of topics which will be used to build the links:

columns = [t[0] for t in frequency] 

Then, we create a list of usernames that are stored as an index in our dataframe:

usernames = users.index.tolist()  

We merge lists of cleaned tokens into a string:

users['clean_join'] = users['clean'].apply(lambda x: " ".join(x))  

For each topic, we create a Boolean variable to check if a topic was expressed by a user and we store this information in a topic column:

 for column in columns: users[column] = users['clean_join'].str.lower().str.contains(column) ...