Create the DataFrames count_df and tfidf_df by using pd.DataFrame() and specifying the values as the first argument and the columns (or features) as the second argument.
- The values can be accessed by using the .A attribute of, respectively, count_train and tfidf_train.
- The columns can be accessed using the .get_feature_names() methods of count_vectorizer and tfidf_vectorizer.
Print the head of each DataFrame to investigate their structure. This has been done for you.
Test if the column names are the same for each DataFrame by creating a new object called difference to see the difference between the columns that count_df has from tfidf_df. Columns can be accessed using the .columns attribute of a DataFrame. Subtract the set of tfidf_df.columns from the set of count_df.columns.
Test if the two DataFrames are equivalent by using the .equals() method on count_df with tfidf_df as the argument.

Training and testing

Introduction to Natural Language Processing in Python