In [1]:

    

%pylab inline


    

    




Populating the interactive namespace from numpy and matplotlib

In [2]:

    

import pandas as pd


    

In [3]:

    

df = pd.read_csv('data/output.tsv', sep="\t")


    

In [4]:

    

locs = df.groupby('pub_location')


    

In [5]:

    

locs_count = locs.count()


    

In [6]:

    

locs_count_mask = locs_count["control_number"] > 500


    

In [7]:

    

top_cities = locs_count[locs_count_mask]


    

In [8]:

    

len(top_cities)


    

    
Out[8]:





891

In [9]:

    

top_cities.to_csv('data/aggs/top_cities.csv')


    

In [10]:

    

years = df.groupby('pub_year')


    

In [11]:

    

year_count = years.count()


    

In [12]:

    

year_count_mask = year_count["control_number"] > 100


    

In [13]:

    

top_years = year_count[year_count_mask]


    

In [14]:

    

len(top_years)


    

    
Out[14]:





1847

In [15]:

    

top_years.to_csv('data/aggs/top_years.csv')


    

In [16]:

    

authors = df.groupby("author")
author_count = authors.count()


    

In [17]:

    

author_count_mask = author_count["control_number"] > 100


    

In [18]:

    

top_authors = author_count[author_count_mask]


    

In [19]:

    

len(top_authors)


    

    
Out[19]:





8072

In [20]:

    

top_authors.to_csv('data/aggs/top_authors.csv')


    

In [21]:

    

texts = df.groupby("title")
text_count = texts.count()


    

In [22]:

    

text_count_mask = text_count["control_number"] > 50


    

In [23]:

    

top_texts = text_count[text_count_mask]


    

In [24]:

    

len(top_texts)


    

    
Out[24]:





3843

In [25]:

    

top_texts.to_csv("data/aggs/top_texts.csv")


    

In [26]:

    

trans = df.groupby("translation")


    

In [27]:

    

trans.count()


    

    
Out[27]:






  

    

      

      
control_number
      
title
      
uniform_title
      
author
      
publisher
      
pub_location
      
pub_year
      
prev_language
    
    

      
translation
      

      

      

      

      

      

      

      

    
  
  

    

      
False
      
12209457
      
12208899
      
379087
      
9371027
      
10875393
      
11008791
      
11284861
      
0
    
    

      
True
      
978788
      
978772
      
477838
      
876342
      
961041
      
961593
      
962432
      
935493
    
  

In [28]:

    

lang = df.groupby(['translation', 'prev_language'])


    

In [29]:

    

lang_count = lang.count()


    

In [30]:

    

mask = lang_count["control_number"] > 5


    

In [31]:

    

top_langs = lang_count[mask]


    

In [32]:

    

len(top_langs)


    

    
Out[32]:





271

In [33]:

    

top_langs.to_csv("data/aggs/top_langs.csv")


    

In [34]:

    

country_year = df.groupby(["pub_location", "pub_year"])


    

In [35]:

    

country_year_count = country_year.count()


    

In [36]:

    

cy_mask = country_year_count["control_number"] > 5


    

In [37]:

    

top_cy = country_year_count[cy_mask]


    

In [38]:

    

len(top_cy)


    

    
Out[38]:





107189

In [39]:

    

top_cy.to_csv("data/aggs/cities_by_year.csv")


    

In [4]:

    

edges = df[["title", "author", "pub_location","pub_year"]]


    

In [5]:

    

edges.to_csv("data/aggs/book_city_edges.tsv", sep="\t")


    

In [6]:

    

locs = df["pub_location"]


    

In [7]:

    

locs.to_csv("data/aggs/book_city_locs.tsv", sep="\t")


    

In [4]:

    

edges = df[["title", "author","pub_year"]]


    

In [6]:

    

locs = pd.read_csv("data/aggs/clean_locs.csv")


    

In [7]:

    

len(edges), len(locs)


    

    
Out[7]:





(13188245, 13188245)

In [11]:

    

new_edges = pd.concat([edges, locs], axis=1)


    

In [14]:

    

new_edges.to_csv("data/aggs/refined_edges.tsv", sep="\t")


    

In [ ]: