In [1]:

    

%load_ext cypher
%matplotlib inline
import pandas as pd


    

    




/home/davebshow/.virtualenvs/scientific3/lib/python3.4/site-packages/IPython/config.py:13: ShimWarning: The `IPython.config` package has been deprecated. You should import from traitlets.config instead.
  "You should import from traitlets.config instead.", ShimWarning)
/home/davebshow/.virtualenvs/scientific3/lib/python3.4/site-packages/IPython/utils/traitlets.py:5: UserWarning: IPython.utils.traitlets has moved to a top-level traitlets package.
  warn("IPython.utils.traitlets has moved to a top-level traitlets package.")

In [5]:

    

top_tweets = %cypher match (n:tweet)-[r]-(m:tweet) return n.text, count(r) as deg order by deg desc limit 10


    

    




10 rows affected.

In [6]:

    

top_tweets.get_dataframe()


    

    
Out[6]:






  

    

      

      
n
      
deg
    
  
  

    

      
0
      
{'text': 'Thoughts go out to everyone in Paris...
      
38199
    
    

      
1
      
{'text': 'thoughts and prayers for those peopl...
      
27827
    
    

      
2
      
{'text': 'Thinking of everyone in Paris.', 'ti...
      
22158
    
    

      
3
      
{'text': 'Had a great first show but just hear...
      
17657
    
    

      
4
      
{'text': 'all of my love and prayers go out to...
      
16860
    
    

      
5
      
{'text': 'To people blaming refugees for attac...
      
14070
    
    

      
6
      
{'text': 'Peace for Paris https://t.co/ryf6XB2...
      
13471
    
    

      
7
      
{'text': 'Paris massacre is an act of terroris...
      
12738
    
    

      
8
      
{'text': 'Please, pray for Paris.', 'tid': '66...
      
12262
    
    

      
9
      
{'text': 'Praying for Paris. Our hearts are br...
      
11636
    
  

In [35]:

    

top_tags = %cypher match (n:hashtag)-[r]-(m) return n.hashtag, count(r) as deg order by deg desc limit 10


    

    




10 rows affected.

In [36]:

    

top_tags.get_dataframe()


    

    
Out[36]:






  

    

      

      
n.hashtag
      
deg
    
  
  

    

      
0
      
paris
      
719571
    
    

      
1
      
prayforparis
      
368931
    
    

      
2
      
parisattacks
      
206159
    
    

      
3
      
prayers4paris
      
52805
    
    

      
4
      
bataclan
      
51524
    
    

      
5
      
rechercheparis
      
46333
    
    

      
6
      
prayersforparis
      
31736
    
    

      
7
      
france
      
31605
    
    

      
8
      
porteouverte
      
30484
    
    

      
9
      
fusillade
      
27833
    
  

In [10]:

    

top_users =  %cypher match (n:user)-[r]-(m) return n.screen_name, count(r) as deg order by deg desc limit 10


    

    




10 rows affected.

In [11]:

    

top_users.get_dataframe()


    

    
Out[11]:






  

    

      

      
n
      
deg
    
  
  

    

      
0
      
{'screen_name': 'Louis_Tomlinson', 'uid': '842...
      
76472
    
    

      
1
      
{'screen_name': 'NiallOfficial', 'uid': '10511...
      
55876
    
    

      
2
      
{'screen_name': 'RecherchesP', 'uid': '4185722...
      
45967
    
    

      
3
      
{'screen_name': 'Harry_Styles', 'uid': '181561...
      
44814
    
    

      
4
      
{'screen_name': 'infos140', 'uid': '1356382759'}
      
41809
    
    

      
5
      
{'screen_name': 'justinbieber', 'uid': '272600...
      
38659
    
    

      
6
      
{'screen_name': 'nytimes', 'uid': '807095'}
      
34432
    
    

      
7
      
{'screen_name': 'Michael5SOS', 'uid': '4032468...
      
33745
    
    

      
8
      
{'screen_name': 'jean_jullien', 'uid': '185122...
      
30970
    
    

      
9
      
{'screen_name': 'AP', 'uid': '51241574'}
      
29869
    
  

In [3]:

    

tweets = %cypher match (n:tweet) where n.lang is not null return n.tid, n.lang


    

    




4263562 rows affected.

In [6]:

    

tweets = tweets.get_dataframe()
tweets.to_csv("data/tweets_w_lang.csv")


    

In [7]:

    

del tweets


    

In [9]:

    

tweet_edges = %cypher match (n:tweet)--(m:tweet) where n.lang is not null and m.lang is not null return n.tid, m.tid


    

    




6090786 rows affected.

In [10]:

    

tweet_edges = tweet_edges.get_dataframe()
tweet_edges.head()


    

    
Out[10]:






  

    

      

      
n.tid
      
m.tid
    
  
  

    

      
0
      
665438816551964672
      
665322019291013120
    
    

      
1
      
665563618776326144
      
665563677429403650
    
    

      
2
      
665496407525208064
      
665496689357254656
    
    

      
3
      
665502309351321602
      
665485168120045568
    
    

      
4
      
665549978979442688
      
665549060665909248
    
  

In [11]:

    

tweet_edges.to_csv("data/tweets_w_lang_edge.csv")


    

In [ ]:

    




    

In [ ]:

    




    

In [13]:

    

tweets_df = pd.read_csv("data/clean_tweets.csv", delimiter="\t")


    

In [15]:

    

tweets_df.head()


    

    
Out[15]:






  

    

      

      
tid:ID
      
lang
      
name
      
text
      
clean_text
      
polarity:float
      
subjectivity:float
      
created_at
      
full_name
      
country
      
country_code
      
coordinates
      
:LABEL
    
  
  

  

In [ ]:

    

lang_groups = tweets_df.groupby("lang")


    

In [ ]:

    




    

In [ ]:

    

geotweets = tweets_df.dropna(subset=["coordinates"])


    

In [ ]:

    

country_groups = geotweets.groupby("country")


    

In [ ]:

    

city_groups = geotweets.groupby("name")


    

In [ ]:

    

french_tweets_df = tweets_df[tweets_df["lang"] == "fr"]


    

In [ ]:

    

english_tweets_df = tweets_df[tweets_df["lang"] == "en"]