In [1]:

    

import pandas as pd
%matplotlib inline
import matplotlib.pyplot as plt # package for doing plotting (necessary for adding the line)
import statsmodels.formula.api as smf # package we'll be using for linear regression
import matplotlib
import pandas as pd
import numpy as np
plt.style.use('ggplot')
import dateutil.parser
import math
import random
import matplotlib.ticker as plticker


matplotlib.rcParams['ps.fonttype'] = 42


    

In [2]:

    

df = pd.read_excel("iran_obama.xlsx")
df.head()


    

    
Out[2]:






  

    

      

      
Iran_Deal
      
Approval_for_Obama
    
  
  

    

      
Democrats
      
60
      
78
    
    

      
Republicans
      
34
      
10
    
    

      
Black
      
54
      
85
    
    

      
White
      
45
      
35
    
    

      
Ind
      
44
      
37
    
  

In [10]:

    

lm = smf.ols(formula="Iran_Deal~Approval_for_Obama",data=df).fit() #notice the formula regresses Y on X (Y~X)
lm.params


    

    
Out[10]:





Intercept             30.730004
Approval_for_Obama     0.354593
dtype: float64

In [11]:

    

intercept, slope = lm.params


    

In [12]:

    

fig, ax = plt.subplots(figsize=(10,5))
df.plot(kind="scatter",x="Approval_for_Obama",y="Iran_Deal", ax=ax)
plt.plot(df["Approval_for_Obama"],slope*df["Approval_for_Obama"]+intercept,"-",color="red")
ax.set_ylim([0, 100])
ax.set_title("Predict Feelings on Iran Deal", color='brown')
ax.set_xlabel('Approval for Obama', color='brown')
ax.set_ylabel('Favour Iran Deal', color='brown')


    

    
Out[12]:





<matplotlib.text.Text at 0x9b28d90>






    

In [6]:

    

lm.summary()


    

    




c:\users\radhika\appdata\local\programs\python\python35-32\lib\site-packages\scipy\stats\stats.py:1535: UserWarning: kurtosistest only valid for n>=20 ... continuing anyway, n=11
  "anyway, n=%i" % int(n))






    
Out[6]:





OLS Regression Results

  
Dep. Variable:
    
Approval_for_Obama
 
  R-squared:         
 
   0.849


  
Model:
                    
OLS
        
  Adj. R-squared:    
 
   0.833


  
Method:
              
Least Squares
   
  F-statistic:       
 
   50.70


  
Date:
              
Tue, 26 Jul 2016
  
  Prob (F-statistic):
 
5.54e-05


  
Time:
                  
23:53:00
      
  Log-Likelihood:    
 
 -38.948


  
No. Observations:
       
    11
       
  AIC:               
 
   81.90


  
Df Residuals:
           
     9
       
  BIC:               
 
   82.69


  
Df Model:
               
     1
       
                     
     
 
   


  
Covariance Type:
       
nonrobust
     
                     
     
 
   




      
         
coef
     
std err
      
t
      
P>|t|
  
[0.025
    
0.975]
  


  
Intercept
 
  -66.5279
 
   16.172
 
   -4.114
 
 0.003
 
 -103.111
 
  -29.945


  
Iran_Deal
 
    2.3950
 
    0.336
 
    7.121
 
 0.000
 
    1.634
 
    3.156




  
Omnibus:
       
11.277
 
  Durbin-Watson:     
 
   2.838


  
Prob(Omnibus):
 
 0.004
 
  Jarque-Bera (JB):  
 
   5.627


  
Skew:
          
 1.476
 
  Prob(JB):          
 
  0.0600


  
Kurtosis:
      
 4.889
 
  Cond. No.          
 
    280.

In [7]:

    

neg_df = pd.read_excel("negiran_obama.xlsx")
neg_df.head()


    

    
Out[7]:






  

    

      

      
Iran_neg
      
Approval_for_Obama
    
  
  

    

      
Democrats
      
29
      
78
    
    

      
Republicans
      
4
      
10
    
    

      
Black
      
37
      
85
    
    

      
White
      
11
      
35
    
    

      
Ind
      
15
      
37
    
  

In [13]:

    

neg_lm = smf.ols(formula="Iran_neg~Approval_for_Obama",data=neg_df).fit() #notice the formula regresses Y on X (Y~X)
neg_lm.params


    

    
Out[13]:





Intercept            -0.762126
Approval_for_Obama    0.399968
dtype: float64

In [14]:

    

intercept, slope = neg_lm.params


    

In [15]:

    

fig, ax = plt.subplots(figsize=(10,5))
neg_df.plot(kind="scatter",x="Approval_for_Obama",y="Iran_neg", ax=ax)
plt.plot(df["Approval_for_Obama"],slope*df["Approval_for_Obama"]+intercept,"-",color="red")
ax.set_ylim([0, 100])
ax.set_title("Opinion on Iran Deal", color='brown')
ax.set_xlabel('Approval for Obama', color='brown')
ax.set_ylabel('Confident in Negotiations with Iran', color='brown')


    

    
Out[15]:





<matplotlib.text.Text at 0x9b3cb90>






    

In [16]:

    

neg_lm.summary()


    

    




c:\users\radhika\appdata\local\programs\python\python35-32\lib\site-packages\scipy\stats\stats.py:1535: UserWarning: kurtosistest only valid for n>=20 ... continuing anyway, n=11
  "anyway, n=%i" % int(n))






    
Out[16]:





OLS Regression Results

  
Dep. Variable:
        
Iran_neg
     
  R-squared:         
 
   0.948


  
Model:
                   
OLS
       
  Adj. R-squared:    
 
   0.942


  
Method:
             
Least Squares
  
  F-statistic:       
 
   163.1


  
Date:
             
Thu, 28 Jul 2016
 
  Prob (F-statistic):
 
4.52e-07


  
Time:
                 
11:18:50
     
  Log-Likelihood:    
 
 -23.340


  
No. Observations:
      
    11
      
  AIC:               
 
   50.68


  
Df Residuals:
          
     9
      
  BIC:               
 
   51.48


  
Df Model:
              
     1
      
                     
     
 
   


  
Covariance Type:
      
nonrobust
    
                     
     
 
   




           
             
coef
     
std err
      
t
      
P>|t|
  
[0.025
    
0.975]
  


  
Intercept
          
   -0.7621
 
    1.616
 
   -0.472
 
 0.648
 
   -4.418
 
    2.893


  
Approval_for_Obama
 
    0.4000
 
    0.031
 
   12.772
 
 0.000
 
    0.329
 
    0.471




  
Omnibus:
       
 0.645
 
  Durbin-Watson:     
 
   2.464


  
Prob(Omnibus):
 
 0.724
 
  Jarque-Bera (JB):  
 
   0.571


  
Skew:
          
 0.153
 
  Prob(JB):          
 
   0.752


  
Kurtosis:
      
 1.927
 
  Cond. No.          
 
    124.

In [ ]: