In [4]:

    

%matplotlib inline

from sklearn.datasets import load_iris
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import precision_recall_fscore_support
from sklearn.linear_model import LinearRegression
import matplotlib.pylab as plt
import pandas as pd
import numpy as np
from pandas.tools.plotting import scatter_matrix


    

In [141]:

    

import json
from pprint import pprint
with open('/Users/danielkershaw/Downloads/part-00004 (5)') as data_file:
    df = pd.read_json(json.load(data_file))


    

In [143]:

    

df.head()


    

    
Out[143]:






  

    

      

      
numberActivatedUsers
      
numberOfActivations
      
period
      
word
    
  
  

    

      
0
      
2
      
3
      
90
      
gainlands
    
    

      
1
      
1
      
2
      
60
      
gainlands
    
    

      
10
      
8
      
14
      
90
      
gagguni
    
    

      
100
      
2
      
14
      
90
      
hayk
    
    

      
1000
      
33
      
60
      
90
      
mansaac
    
  

In [146]:

    

scatter_matrix(df.pivot(index='word', columns='period', values='numberOfActivations'), alpha=0.2, figsize=(15, 15), diagonal='kde')


    

    
Out[146]:





array([[<matplotlib.axes._subplots.AxesSubplot object at 0x11afb5710>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x11939ff90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x113221110>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x110d3f9d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x11586ae50>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x119a3ba90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x11a46f990>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x113b87d10>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1150532d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x118b1e150>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x114dcced0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1149980d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x117352f10>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x11631ded0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x11e573d50>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x118e17c90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1153b7b10>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x118e45b90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x113ab99d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x113c5f850>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x114910210>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x118687090>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x1141c9090>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x118c57050>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x114ce0150>]], dtype=object)






    

In [84]:

    

# df[(df["popularity"]< 1)].pivot(index='word', columns='timedelta', values='popularity').plot(kind='scatter',x=90, y=7)
df.pivot(index='word', columns='period', values='activation').plot(kind='scatter',x=7, y=30, loglog=True)


    

    
Out[84]:





<matplotlib.axes._subplots.AxesSubplot at 0x10f838510>






    

In [87]:

    




    

    
Out[87]:





<matplotlib.axes._subplots.AxesSubplot at 0x10fd6f090>






    

In [145]:

    

import statsmodels.api as sm
import statsmodels.formula.api as smf
dfp = df.pivot(index='word', columns='period', values='numberActivatedUsers')
ttt = [7,14,30,60,90]
dfp.columns = ['a', 'b', 'c', 'd', 'e']

fig, axes = plt.subplots(nrows=5, ncols=5, figsize=(25, 25))

for x1 in range(0, len(dfp.columns)):
    for y1 in range(0, len(dfp.columns)):
        results =  smf.ols(dfp.columns[y1]+' ~ '+dfp.columns[x1], dfp).fit()
        intercept, slope = results.params
        r2 = results.rsquared
        x = np.array([min(dfp[dfp.columns[x1]]), max(dfp[dfp.columns[x1]])])
        y = intercept + slope * x
        ax = df.pivot(index='word', columns='period', values='numberActivatedUsers').plot(kind='scatter',x=ttt[x1], y=ttt[y1], loglog=True, ax=axes[x1,y1])
        axes[x1,y1].plot(x, y, 'r-')
        x_lims = axes[x1,y1].get_xlim()
        y_lims = axes[x1,y1].get_ylim()
        axes[x1,y1].plot(y_lims, x_lims, 'g-')


    

In [ ]: