In [6]:

    

import json
import numpy as np
import pandas as pd
import datetime
import os
import matplotlib
import matplotlib.pyplot as plt
import seaborn as sns
colours = sns.color_palette()
sns.set_style("white")
sns.set_context("poster")

%matplotlib inline


    

In [7]:

    

columns = ['topic1', 'topic2', 'topic3', 'time']

data_df = pd.DataFrame(columns=columns)
data_df.index.name = 'respondent'
data_df


    

    
Out[7]:






  

    

      

      
topic1
      
topic2
      
topic3
      
time
    
    

      
respondent
      

      

      

      

    
  
  

  

In [8]:

    

data_path = './../responses/'
# Deal with question 1

i = 0
for index, file_ in enumerate(os.listdir(data_path)):
    
    if ('second_stage_' in file_):
       
        json_data=open(os.path.join(data_path, file_)).read()
        data = json.loads(json_data)

        for selection in data['choices']:
                        
            data_df.set_value(i, selection['question'].replace(' ', '').lower(), int(selection['selected_id']))
        
        data_df.set_value(i, 'time', datetime.datetime.fromtimestamp(data['time']/1000))
        i += 1

data_df


    

    
Out[8]:






  

    

      

      
topic1
      
topic2
      
topic3
      
time
    
    

      
respondent
      

      

      

      

    
  
  

    

      
0
      
2
      
3
      
2
      
2017-05-02 09:30:17
    
    

      
1
      
3
      
3
      
3
      
2017-05-02 09:41:22
    
    

      
2
      
2
      
2
      
3
      
2017-05-02 09:41:36
    
    

      
3
      
3
      
3
      
3
      
2017-05-02 10:41:57
    
    

      
4
      
2
      
1
      
3
      
2017-05-02 12:29:10
    
    

      
5
      
3
      
3
      
3
      
2017-05-02 13:21:30
    
    

      
6
      
3
      
1
      
1
      
2017-05-02 14:22:47
    
    

      
7
      
1
      
3
      
1
      
2017-05-02 14:23:08
    
    

      
8
      
2
      
2
      
1
      
2017-05-02 14:23:37
    
    

      
9
      
3
      
1
      
3
      
2017-05-02 14:27:12
    
    

      
10
      
1
      
3
      
2
      
2017-05-03 13:24:50
    
    

      
11
      
1
      
2
      
3
      
2017-05-03 21:51:19
    
    

      
12
      
2
      
1
      
3
      
2017-05-04 03:47:56
    
    

      
13
      
2
      
2
      
3
      
2017-05-04 08:45:34
    
    

      
14
      
3
      
3
      
1
      
2017-05-04 09:34:22
    
    

      
15
      
1
      
2
      
2
      
2017-05-04 15:57:33
    
    

      
16
      
1
      
3
      
3
      
2017-05-05 11:49:37
    
  

In [9]:

    

vals = [data_df['topic{}'.format(i)].values for i in range(1,4)]


    

In [11]:

    

plt.figure(figsize=(5,4))
plt.hist(vals, label=['Topic 1', 'Topic 2', 'Topic 3'], alpha=0.8)
plt.legend(loc='best')
plt.xticks([1.1,2.1,2.9], ['Ranking', 'Cardinal Score', 'Plurality'])
plt.ylabel('count')
plt.show()


    

In [ ]:

    




    

In [ ]:

    




    

In [ ]: