notebook.community

Edit and run

Assignment 4

Using data from this FiveThirtyEight post, write code to calculate the correlation of the responses from the poll. Respond to the story in your PR. Is this a good example of data journalism? Why or why not?

Extracted polling info on pages 20/21 using Tabula. (Astoundingly good results, if you leave the headers out)





In [156]:



    


import pandas as pd
%matplotlib inline
import matplotlib.pyplot as plt
import statsmodels.formula.api as smf

First for the Obama Love and Iran Deal Approval





In [157]:



    


df_Obama = pd.read_csv("data/Fox_polls - Obama Job.csv")



    











In [158]:



    


df_Iran_Deal = pd.read_csv("data/Fox_polls - Iran Deal.csv")



    











In [159]:



    


df_Iran_Nego = pd.read_csv("data/Fox_polls - Iran Nego.csv")



    











In [160]:



    


df_Obama.head(3)



    


















    
Out[160]:










  
    

      
      Unnamed: 0
      Approve
      Disapprove
      (Don't know)
    

  
  
    

      0
      Dem
      0.78
      0.17
      0.05
    

    

      1
      Rep
      0.10
      0.86
      0.04
    

    

      2
      Ind
      0.37
      0.53
      0.10
    

  




















In [161]:



    


df_Iran_Deal.head(3)



    


















    
Out[161]:










  
    

      
      Unnamed: 0
      Favor
      Oppose
      (Don't know)
    

  
  
    

      0
      Dem
      0.60
      0.26
      0.14
    

    

      1
      Rep
      0.34
      0.60
      0.06
    

    

      2
      Ind
      0.44
      0.49
      0.07
    

  




















In [162]:



    


df_Obama_Iran_Deal = df_Obama.merge(df_Iran_Deal, left_on = 'Unnamed: 0', right_on='Unnamed: 0')



    











In [163]:



    


del df_Obama_Iran_Deal['Disapprove']
del df_Obama_Iran_Deal["(Don't know)_x"]
del df_Obama_Iran_Deal["Oppose"]
del df_Obama_Iran_Deal["(Don't know)_y"]
df_Obama_Iran_Deal.head(3)



    


















    
Out[163]:










  
    

      
      Unnamed: 0
      Approve
      Favor
    

  
  
    

      0
      Dem
      0.78
      0.60
    

    

      1
      Rep
      0.10
      0.34
    

    

      2
      Ind
      0.37
      0.44
    

  




















In [164]:



    


df_Obama_Iran_Deal.columns = ['Group', 'Obama', 'Iran_Deal']



    











In [165]:



    


fig, ax = plt.subplots(figsize =(7,5))

#Font
csfont = {'fontname':'DIN Condensed'}

lm = smf.ols(formula='Iran_Deal~Obama',data=df_Obama_Iran_Deal).fit()
lm.params
Intercept, Obama_love = lm.params

df_Obama_Iran_Deal.plot(kind='scatter', x='Obama', y='Iran_Deal', ax= ax, color='tomato')
ax.plot(df_Obama_Iran_Deal["Obama"],Obama_love*df_Obama_Iran_Deal["Obama"]+Intercept,"-",color="green")

ax.set_axis_bgcolor("WhiteSmoke")

ax.set_ylabel('')

ax.xaxis.grid(color='darkgrey', linestyle=':', linewidth=0.5)
ax.yaxis.grid(color='darkgrey', linestyle=':', linewidth=0.5)

ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.spines['bottom'].set_visible(False)
ax.spines['left'].set_visible(False)

plt.tick_params(
    #axis='x',
    top='off',
    which='off',
    left='off',
    right='off',
    bottom='off',
    labeltop='off',
    labelbottom='off')

#labelling, getting rid of boarders
ax.set_xlabel('Obama Love', **csfont, fontsize=12)
ax.set_title("Obama Love versus approval of Iran Deal", **csfont, fontsize=24)
ax.set_ylabel('Iran Deal Approval', **csfont, fontsize=12)

ax.set_axisbelow(True)

Now for Obama Love and Confidence in Negotiations with Iran





In [166]:



    


df_Obama.head()



    


















    
Out[166]:










  
    

      
      Unnamed: 0
      Approve
      Disapprove
      (Don't know)
    

  
  
    

      0
      Dem
      0.78
      0.17
      0.05
    

    

      1
      Rep
      0.10
      0.86
      0.04
    

    

      2
      Ind
      0.37
      0.53
      0.10
    

    

      3
      Men
      0.41
      0.53
      0.05
    

    

      4
      Women
      0.47
      0.47
      0.06
    

  




















In [167]:



    


df_Iran_Nego.head()



    


















    
Out[167]:










  
    

      
      Unnamed: 0
      Very confident
      Somewhat confident
      Not very confident
      Not at all confident
      (Don't know)
    

  
  
    

      0
      Dem
      0.29
      0.49
      0.11
      0.08
      0.04
    

    

      1
      Rep
      0.04
      0.13
      0.26
      0.57
      0.01
    

    

      2
      Ind
      0.15
      0.29
      0.22
      0.32
      0.02
    

    

      3
      Men
      0.17
      0.28
      0.18
      0.35
      0.02
    

    

      4
      Women
      0.17
      0.35
      0.19
      0.27
      0.02
    

  




















In [168]:



    


df_Iran_Nego['Confident'] = df_Iran_Nego['Very confident'] + df_Iran_Nego['Somewhat confident']



    











In [169]:



    


df_Obama_Iran_Nego = df_Obama.merge(df_Iran_Nego, left_on = 'Unnamed: 0', right_on='Unnamed: 0')



    











In [170]:



    


del df_Obama_Iran_Nego['Disapprove']
del df_Obama_Iran_Nego["(Don't know)_x"]
del df_Obama_Iran_Nego["Very confident"]
del df_Obama_Iran_Nego["Somewhat confident"]
del df_Obama_Iran_Nego['Not very confident']
del df_Obama_Iran_Nego["Not at all confident"]
del df_Obama_Iran_Nego["(Don't know)_y"]
df_Obama_Iran_Nego.head()



    


















    
Out[170]:










  
    

      
      Unnamed: 0
      Approve
      Confident
    

  
  
    

      0
      Dem
      0.78
      0.78
    

    

      1
      Rep
      0.10
      0.17
    

    

      2
      Ind
      0.37
      0.44
    

    

      3
      Men
      0.41
      0.45
    

    

      4
      Women
      0.47
      0.52
    

    

      5
      White
      0.35
      0.40
    

    

      6
      Black
      0.85
      0.83
    

    

      7
      College Degree
      0.47
      0.55
    

    

      8
      No Degree
      0.43
      0.43
    

    

      9
      under 35
      0.56
      0.57
    

    

      10
      35-54
      0.35
      0.42
    

    

      11
      55+
      0.45
      0.48
    

    

      12
      65+
      0.43
      0.44
    

    

      13
      under $50k
      0.48
      0.52
    

    

      14
      $50k+
      0.42
      0.47
    

    

      15
      Lib
      0.69
      0.72
    

    

      16
      Cons
      0.24
      0.29
    

    

      17
      Tea Party
      0.15
      0.17
    

  




















In [177]:



    


df_Obama_Iran_Nego.columns = ['Group', 'ObamaApp', 'Confidence']



    











In [178]:



    


df_Obama_Iran_Nego.head()



    


















    
Out[178]:










  
    

      
      Group
      ObamaApp
      Confidence
    

  
  
    

      0
      Dem
      0.78
      0.78
    

    

      1
      Rep
      0.10
      0.17
    

    

      2
      Ind
      0.37
      0.44
    

    

      3
      Men
      0.41
      0.45
    

    

      4
      Women
      0.47
      0.52
    

  




















In [179]:



    


fig, ax = plt.subplots(figsize =(7,5))

#Font
csfont = {'fontname':'DIN Condensed'}

lm = smf.ols(formula='Confidence~ObamaApp',data=df_Obama_Iran_Nego).fit()
lm.params
Intercept, Obama_love = lm.params

df_Obama_Iran_Nego.plot(kind='scatter', x='ObamaApp', y='ObamaApp', ax= ax, color='tomato')
ax.plot(df_Obama_Iran_Nego["ObamaApp"],Obama_love*df_Obama_Iran_Nego["ObamaApp"]+Intercept,"-",color="green")

ax.set_axis_bgcolor("WhiteSmoke")

ax.set_ylabel('')

ax.xaxis.grid(color='darkgrey', linestyle=':', linewidth=0.5)
ax.yaxis.grid(color='darkgrey', linestyle=':', linewidth=0.5)

ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.spines['bottom'].set_visible(False)
ax.spines['left'].set_visible(False)

plt.tick_params(
    #axis='x',
    top='off',
    which='off',
    left='off',
    right='off',
    bottom='off',
    labeltop='off',
    labelbottom='off')

#labelling, getting rid of boarders
ax.set_xlabel('Obama Love', **csfont, fontsize=12)
ax.set_title("Obama Love versus confidence in Admins Negotiations", **csfont, fontsize=24)
ax.set_ylabel('Confindence in Negotiations', **csfont, fontsize=12)

ax.set_axisbelow(True)



    


















    
























In [ ]:

Content source: ledeprogram/algorithms

Similar notebooks:

notebook.community | gallery | about