notebook.community

Edit and run 





In [1]:



    


import pandas as pd
import numpy as np
import pickle
from datetime import date
from collections import Counter
from pandas import Series
from patsy import dmatrices, dmatrix
from collections import Counter

Loading Data





In [2]:



    


fh = open("/data/csc/fb_persons/unified_100percentpeople.pkl", 'rb')
df = pickle.load(fh)



    











In [3]:



    


len(df)



    


















    
Out[3]:








3379996

















In [4]:



    


df.head()



    


















    
Out[4]:










  
    

      
      date_of_birth
      ethnicity
      gender
      name
      nationality
      place_of_birth
      profession
      religion
    

  
  
    

      0
              NaN
       None
       Female
       Courtney Jamieson
                 None
          NaN
                  music_art
       None
    

    

      1
              NaN
       None
         Male
             Robert Moir
                 None
          NaN
                      other
       None
    

    

      2
       1962-04-21
       None
         Male
             Mehdi Jomaa
       middle eastern
       Mahdia
       (law_politics, stem)
       None
    

    

      3
              NaN
       None
         Male
        Victoria Shields
                 None
          NaN
                       None
       None
    

    

      4
       1992-06-24
       None
         Male
         Terrick Colston
                 None
          NaN
                       None
       None
    

  




















In [5]:



    


selected_columns = ["gender", "date_of_birth", "ethnicity", "name", "profession",
                    "religion", "nationality"]
df = df[selected_columns]



    











In [6]:



    


df.head()



    


















    
Out[6]:










  
    

      
      gender
      date_of_birth
      ethnicity
      name
      profession
      religion
      nationality
    

  
  
    

      0
       Female
              NaN
       None
       Courtney Jamieson
                  music_art
       None
                 None
    

    

      1
         Male
              NaN
       None
             Robert Moir
                      other
       None
                 None
    

    

      2
         Male
       1962-04-21
       None
             Mehdi Jomaa
       (law_politics, stem)
       None
       middle eastern
    

    

      3
         Male
              NaN
       None
        Victoria Shields
                       None
       None
                 None
    

    

      4
         Male
       1992-06-24
       None
         Terrick Colston
                       None
       None
                 None

Considering only people after the 1800s





In [7]:



    


victorian_age = date(1800, 1, 1)
df = df[np.logical_not(df.date_of_birth.isnull().values)]
df = df[df.date_of_birth > victorian_age]



    











In [8]:



    


len(df)



    


















    
Out[8]:








1300159

















In [9]:



    


len(df.nationality.dropna())



    


















    
Out[9]:








703447

Considering Only Americans





In [10]:



    


americans = map(lambda x: (isinstance(x, basestring) and x == 'united states of america') or
                          (isinstance(x, tuple) and  'united states of america' in x),
                          df.nationality)
df = df[americans]



    











In [11]:



    


len(df)



    


















    
Out[11]:








223632

















In [12]:



    


df.head()



    


















    
Out[12]:










  
    

      
      gender
      date_of_birth
      ethnicity
      name
      profession
      religion
      nationality
    

  
  
    

      11
       Male
       1995-04-09
       None
              Justin Cook
       None
       None
       united states of america
    

    

      44
       Male
       1991-08-29
       None
       Brandon Cunningham
       None
       None
       united states of america
    

    

      47
       Male
       1992-11-06
       None
               Mike Davis
       None
       None
       united states of america
    

    

      48
       Male
       1994-04-15
       None
              Ricky Ortiz
       None
       None
       united states of america
    

    

      53
       Male
       1993-06-03
       None
            Jamal Douglas
       None
       None
       united states of america

Stats Models





In [13]:



    


import statsmodels as sm
from statsmodels.discrete.discrete_model import MNLogit
from statsmodels.tools import add_constant

Drop 'Other' people





In [14]:



    


other_index = np.where((df.gender == 'Other').values)[0]
df.drop(df.index[other_index], inplace=True)



    











In [15]:



    


def replace(x):
  if not x:
    return np.nan
  if isinstance(x, tuple):
    return "mixed"
  return x



    











In [16]:



    


df["religion"] = df.religion.apply(func=lambda x: replace(x))



    











In [17]:



    


def replace_nationality(x):
  if not x: return np.nan
  if x == "united states of america":
    return "USA"
  if isinstance(x, tuple):
    nationalities = set(x)
    x = nationalities.difference(["united states of america"])
    if not x: return "USA"
    if len(x) > 1: return "mixed"
    x = list(x)[0]
  return x



    











In [18]:



    


df["nationality"] = df.nationality.apply(func=lambda x: replace_nationality(x))

Drop People without profession





In [19]:



    


empty_rows = np.where([x is None for x in df.profession.values.flatten()])[0]
dfXy = df.drop(df.index[empty_rows])
dfXy = dfXy[["profession", "gender", "religion", "nationality"]]

Map Professions to traditional vs non-traditional





In [20]:



    


prof_map = {'music_art': "traditional", 'humanities': "traditional",
            'athlete': "non-traditional", 'business': "non-traditional",
             'law_politics': "non-traditional", 'religion': "non-traditional",
             'stem': "non-traditional"}



    











In [21]:



    


def clear_prof(x):
  if isinstance(x, basestring):
    return prof_map.get(x, x)
  else:
    tmp = set([prof_map.get(x_, x_) for x_ in x])
    if len(tmp) == 1:
      return list(tmp)[0]
    else:
      return "mixed"



    











In [22]:



    


dfXy["profession"] = dfXy.profession.apply(func=lambda x: clear_prof(x))

Save Results





In [28]:



    


dfXy.nationality.value_counts()



    


















    
Out[28]:








USA                 154456
white                 5841
asian                  828
hispanic_latino        738
middle eastern         461
african_american       302
mixed                  112
indigenous              45
dtype: int64

















In [26]:



    


dfXy.to_csv("Profession_Analysis_Processed.csv")



    











In [67]:



    


X = dmatrix("C(gender, Treatment('Male')) + "
            "C(nationality, Treatment('USA')) + "
            "C(religion, Treatment('christian'))",
            data=dfXy.fillna("missing"), return_type='dataframe')



    











In [68]:



    


len(X)



    


















    
Out[68]:








162783

















In [69]:



    


X2  = X[[x  for x in X.columns if 'missing' not in x]]



    











In [70]:



    


X.head()



    


















    
Out[70]:










  
    

      
      Intercept
      C(gender, Treatment('Male'))[T.Female]
      C(gender, Treatment('Male'))[T.missing]
      C(nationality, Treatment('USA'))[T.african_american]
      C(nationality, Treatment('USA'))[T.asian]
      C(nationality, Treatment('USA'))[T.hispanic_latino]
      C(nationality, Treatment('USA'))[T.indigenous]
      C(nationality, Treatment('USA'))[T.middle eastern]
      C(nationality, Treatment('USA'))[T.mixed]
      C(nationality, Treatment('USA'))[T.white]
      C(religion, Treatment('christian'))[T.asian religion]
      C(religion, Treatment('christian'))[T.atheist_spiritual]
      C(religion, Treatment('christian'))[T.buddhist]
      C(religion, Treatment('christian'))[T.hindu]
      C(religion, Treatment('christian'))[T.jewish]
      C(religion, Treatment('christian'))[T.missing]
      C(religion, Treatment('christian'))[T.mixed]
      C(religion, Treatment('christian'))[T.muslim]
      C(religion, Treatment('christian'))[T.other]
    

  
  
    

      74 
       1
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       1
       0
       0
       0
    

    

      75 
       1
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       1
       0
       0
       0
    

    

      149
       1
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       1
       0
       0
       0
    

    

      228
       1
       0
       0
       0
       0
       0
       0
       0
       0
       1
       0
       0
       0
       0
       0
       1
       0
       0
       0
    

    

      377
       1
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       0
       1
       0
       0
       0
    

  




















In [71]:



    


y = dfXy["profession"]
ynames = {'mixed': 2, 'non-traditional': 1, 'other': 3, 'traditional': 0}
ynames_reverse = {v:k for k,v in ynames.items()}
y1 = y.apply(lambda x: ynames[x])



    











In [1]:



    


y.head()



    


















    






---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-1-073ab239a5bf> in <module>()
----> 1 y.head()

NameError: name 'y' is not defined
















In [73]:



    


dict(Counter(y.values))



    


















    
Out[73]:








{'mixed': 11514,
 'non-traditional': 48507,
 'other': 2443,
 'traditional': 100319}

















In [74]:



    


print "Colinearity before removing the missing value", np.linalg.cond(X)
print "Colinearity after removing the missing value", np.linalg.cond(X2)



    


















    






Colinearity before removing the missing value 113.900214164
Colinearity after removing the missing value 83.2171475383

















In [75]:



    


#data = dfX.values
#data = add_constant(dfX2)
model = MNLogit(y1, X2)
result = model.fit(method='ncg', maxiter=100)



    


















    






Optimization terminated successfully.
         Current function value: 0.873872
         Iterations: 89
         Function evaluations: 90
         Gradient evaluations: 178
         Hessian evaluations: 89

















In [76]:



    


result.summary(yname_list=[ynames_reverse[i] for i in range(1,len(ynames))])



    


















    
Out[76]:








MNLogit Regression Results


  Dep. Variable:    profession      No. Observations:     162783   




  Model:              MNLogit       Df Residuals:         162732   




  Method:               MLE         Df Model:                 48   




  Date:          Mon, 02 Feb 2015   Pseudo R-squ.:        0.03914  




  Time:              01:01:24       Log-Likelihood:     -1.4225e+05




  converged:           True         LL-Null:            -1.4805e+05




                                    LLR p-value:           0.000   








                       non-traditional                        coef     std err      z      P>|z| [95.0% Conf. Int.] 




  Intercept                                                   -0.4522     0.006   -72.912  0.000    -0.464    -0.440




  C(gender, Treatment('Male'))[T.Female]                      -1.6275     0.018   -88.625  0.000    -1.664    -1.592




  C(nationality, Treatment('USA'))[T.african_american]         0.5413     0.131     4.148  0.000     0.285     0.797




  C(nationality, Treatment('USA'))[T.asian]                   -0.4445     0.095    -4.690  0.000    -0.630    -0.259




  C(nationality, Treatment('USA'))[T.hispanic_latino]         -0.0328     0.087    -0.379  0.705    -0.203     0.137




  C(nationality, Treatment('USA'))[T.indigenous]               2.1484     0.424     5.065  0.000     1.317     2.980




  C(nationality, Treatment('USA'))[T.middle eastern]          -0.2090     0.116    -1.798  0.072    -0.437     0.019




  C(nationality, Treatment('USA'))[T.mixed]                    0.1930     0.218     0.883  0.377    -0.235     0.621




  C(nationality, Treatment('USA'))[T.white]                   -0.0838     0.032    -2.643  0.008    -0.146    -0.022




  C(religion, Treatment('christian'))[T.asian religion]        0.8817     0.451     1.954  0.051    -0.003     1.766




  C(religion, Treatment('christian'))[T.atheist_spiritual]    -0.2715     0.130    -2.081  0.037    -0.527    -0.016




  C(religion, Treatment('christian'))[T.buddhist]             -1.4632     0.354    -4.133  0.000    -2.157    -0.769




  C(religion, Treatment('christian'))[T.hindu]                 0.6567     0.211     3.112  0.002     0.243     1.070




  C(religion, Treatment('christian'))[T.jewish]                0.6840     0.068    10.125  0.000     0.552     0.816




  C(religion, Treatment('christian'))[T.mixed]                -0.1477     0.229    -0.645  0.519    -0.596     0.301




  C(religion, Treatment('christian'))[T.muslim]               -0.6471     0.191    -3.383  0.001    -1.022    -0.272




  C(religion, Treatment('christian'))[T.other]                 0.2713     0.179     1.518  0.129    -0.079     0.621




                            mixed                             coef     std err      z      P>|z| [95.0% Conf. Int.] 




  Intercept                                                   -2.0706     0.011  -180.641  0.000    -2.093    -2.048




  C(gender, Treatment('Male'))[T.Female]                      -0.6307     0.025   -25.185  0.000    -0.680    -0.582




  C(nationality, Treatment('USA'))[T.african_american]         1.0090     0.175     5.757  0.000     0.665     1.353




  C(nationality, Treatment('USA'))[T.asian]                    0.5254     0.110     4.773  0.000     0.310     0.741




  C(nationality, Treatment('USA'))[T.hispanic_latino]          0.2543     0.133     1.909  0.056    -0.007     0.515




  C(nationality, Treatment('USA'))[T.indigenous]               1.5635     0.629     2.484  0.013     0.330     2.797




  C(nationality, Treatment('USA'))[T.middle eastern]           0.6410     0.140     4.593  0.000     0.367     0.915




  C(nationality, Treatment('USA'))[T.mixed]                    0.4499     0.320     1.407  0.160    -0.177     1.077




  C(nationality, Treatment('USA'))[T.white]                    0.3727     0.046     8.166  0.000     0.283     0.462




  C(religion, Treatment('christian'))[T.asian religion]        0.7351     0.658     1.117  0.264    -0.555     2.025




  C(religion, Treatment('christian'))[T.atheist_spiritual]     1.3621     0.123    11.084  0.000     1.121     1.603




  C(religion, Treatment('christian'))[T.buddhist]              0.8253     0.240     3.436  0.001     0.355     1.296




  C(religion, Treatment('christian'))[T.hindu]                 0.7764     0.282     2.757  0.006     0.225     1.328




  C(religion, Treatment('christian'))[T.jewish]                1.3252     0.082    16.186  0.000     1.165     1.486




  C(religion, Treatment('christian'))[T.mixed]                 1.9005     0.183    10.379  0.000     1.542     2.259




  C(religion, Treatment('christian'))[T.muslim]                0.4177     0.218     1.914  0.056    -0.010     0.845




  C(religion, Treatment('christian'))[T.other]                 1.2020     0.202     5.948  0.000     0.806     1.598




                            other                             coef     std err      z      P>|z| [95.0% Conf. Int.] 




  Intercept                                                   -3.5629     0.023  -153.305  0.000    -3.608    -3.517




  C(gender, Treatment('Male'))[T.Female]                      -0.6510     0.053   -12.284  0.000    -0.755    -0.547




  C(nationality, Treatment('USA'))[T.african_american]        -4.4058     4.835    -0.911  0.362   -13.883     5.071




  C(nationality, Treatment('USA'))[T.asian]                   -0.6227     0.383    -1.625  0.104    -1.374     0.128




  C(nationality, Treatment('USA'))[T.hispanic_latino]         -0.5996     0.412    -1.457  0.145    -1.406     0.207




  C(nationality, Treatment('USA'))[T.indigenous]               1.6349     1.073     1.524  0.127    -0.467     3.737




  C(nationality, Treatment('USA'))[T.middle eastern]          -0.6392     0.507    -1.261  0.207    -1.633     0.354




  C(nationality, Treatment('USA'))[T.mixed]                   -0.3855     1.009    -0.382  0.702    -2.363     1.592




  C(nationality, Treatment('USA'))[T.white]                    0.1094     0.106     1.028  0.304    -0.099     0.318




  C(religion, Treatment('christian'))[T.asian religion]       -1.7469     5.172    -0.338  0.736   -11.883     8.389




  C(religion, Treatment('christian'))[T.atheist_spiritual]     0.1021     0.415     0.246  0.805    -0.710     0.915




  C(religion, Treatment('christian'))[T.buddhist]             -0.7045     1.006    -0.700  0.484    -2.677     1.268




  C(religion, Treatment('christian'))[T.hindu]                -0.1409     1.013    -0.139  0.889    -2.126     1.844




  C(religion, Treatment('christian'))[T.jewish]               -0.4215     0.358    -1.178  0.239    -1.123     0.280




  C(religion, Treatment('christian'))[T.mixed]                -0.5068     1.008    -0.503  0.615    -2.483     1.469




  C(religion, Treatment('christian'))[T.muslim]                1.1908     0.332     3.591  0.000     0.541     1.841




  C(religion, Treatment('christian'))[T.other]                 1.7997     0.298     6.043  0.000     1.216     2.383




















In [77]:



    


print "number of labels is", model.J, "number of features is", model.K



    


















    






number of labels is 4.0 number of features is 17.0

Parameters





In [78]:



    


for i in range(1,len(ynames)):
  print i-1, ":", ynames_reverse[i]



    


















    






0 : non-traditional
1 : mixed
2 : other

















In [79]:



    


params = result.params
params



    


















    
Out[79]:










  
    

      
      0
      1
      2
    

  
  
    

      Intercept
      -0.452178
      -2.070631
      -3.562928
    

    

      C(gender, Treatment('Male'))[T.Female]
      -1.627548
      -0.630670
      -0.650982
    

    

      C(nationality, Treatment('USA'))[T.african_american]
       0.541280
       1.009002
      -4.405795
    

    

      C(nationality, Treatment('USA'))[T.asian]
      -0.444467
       0.525377
      -0.622702
    

    

      C(nationality, Treatment('USA'))[T.hispanic_latino]
      -0.032839
       0.254258
      -0.599552
    

    

      C(nationality, Treatment('USA'))[T.indigenous]
       2.148423
       1.563519
       1.634937
    

    

      C(nationality, Treatment('USA'))[T.middle eastern]
      -0.208955
       0.640987
      -0.639186
    

    

      C(nationality, Treatment('USA'))[T.mixed]
       0.192999
       0.449902
      -0.385529
    

    

      C(nationality, Treatment('USA'))[T.white]
      -0.083787
       0.372693
       0.109439
    

    

      C(religion, Treatment('christian'))[T.asian religion]
       0.881683
       0.735074
      -1.746942
    

    

      C(religion, Treatment('christian'))[T.atheist_spiritual]
      -0.271548
       1.362085
       0.102135
    

    

      C(religion, Treatment('christian'))[T.buddhist]
      -1.463246
       0.825275
      -0.704541
    

    

      C(religion, Treatment('christian'))[T.hindu]
       0.656731
       0.776449
      -0.140925
    

    

      C(religion, Treatment('christian'))[T.jewish]
       0.683999
       1.325236
      -0.421531
    

    

      C(religion, Treatment('christian'))[T.mixed]
      -0.147686
       1.900456
      -0.506792
    

    

      C(religion, Treatment('christian'))[T.muslim]
      -0.647086
       0.417721
       1.190757
    

    

      C(religion, Treatment('christian'))[T.other]
       0.271282
       1.202023
       1.799701

Odd ratios





In [80]:



    


for i in range(1,len(ynames)):
  print i-1, ":", ynames_reverse[i]



    


















    






0 : non-traditional
1 : mixed
2 : other

















In [81]:



    


odds = np.exp(result.params)
odds



    


















    
Out[81]:










  
    

      
      0
      1
      2
    

  
  
    

      Intercept
       0.636241
       0.126106
       0.028356
    

    

      C(gender, Treatment('Male'))[T.Female]
       0.196411
       0.532235
       0.521533
    

    

      C(nationality, Treatment('USA'))[T.african_american]
       1.718205
       2.742861
       0.012206
    

    

      C(nationality, Treatment('USA'))[T.asian]
       0.641166
       1.691096
       0.536493
    

    

      C(nationality, Treatment('USA'))[T.hispanic_latino]
       0.967695
       1.289504
       0.549057
    

    

      C(nationality, Treatment('USA'))[T.indigenous]
       8.571330
       4.775597
       5.129132
    

    

      C(nationality, Treatment('USA'))[T.middle eastern]
       0.811431
       1.898354
       0.527722
    

    

      C(nationality, Treatment('USA'))[T.mixed]
       1.212882
       1.568158
       0.680091
    

    

      C(nationality, Treatment('USA'))[T.white]
       0.919627
       1.451638
       1.115653
    

    

      C(religion, Treatment('christian'))[T.asian religion]
       2.414960
       2.085636
       0.174306
    

    

      C(religion, Treatment('christian'))[T.atheist_spiritual]
       0.762198
       3.904325
       1.107533
    

    

      C(religion, Treatment('christian'))[T.buddhist]
       0.231484
       2.282509
       0.494335
    

    

      C(religion, Treatment('christian'))[T.hindu]
       1.928478
       2.173740
       0.868555
    

    

      C(religion, Treatment('christian'))[T.jewish]
       1.981787
       3.763073
       0.656042
    

    

      C(religion, Treatment('christian'))[T.mixed]
       0.862702
       6.688946
       0.602425
    

    

      C(religion, Treatment('christian'))[T.muslim]
       0.523569
       1.518498
       3.289570
    

    

      C(religion, Treatment('christian'))[T.other]
       1.311644
       3.326841
       6.047837

p-values





In [82]:



    


result.pvalues



    


















    
Out[82]:










  
    

      
      0
      1
      2
    

  
  
    

      Intercept
       0.000000e+00
        0.000000e+00
       0.000000e+00
    

    

      C(gender, Treatment('Male'))[T.Female]
       0.000000e+00
       5.850423e-140
       1.105136e-34
    

    

      C(nationality, Treatment('USA'))[T.african_american]
       3.360756e-05
        8.567901e-09
       3.622180e-01
    

    

      C(nationality, Treatment('USA'))[T.asian]
       2.731140e-06
        1.819300e-06
       1.040626e-01
    

    

      C(nationality, Treatment('USA'))[T.hispanic_latino]
       7.050353e-01
        5.628118e-02
       1.452454e-01
    

    

      C(nationality, Treatment('USA'))[T.indigenous]
       4.080337e-07
        1.298575e-02
       1.274512e-01
    

    

      C(nationality, Treatment('USA'))[T.middle eastern]
       7.212110e-02
        4.371501e-06
       2.073249e-01
    

    

      C(nationality, Treatment('USA'))[T.mixed]
       3.770247e-01
        1.595320e-01
       7.023441e-01
    

    

      C(nationality, Treatment('USA'))[T.white]
       8.229002e-03
        3.198316e-16
       3.039058e-01
    

    

      C(religion, Treatment('christian'))[T.asian religion]
       5.068633e-02
        2.641521e-01
       7.355206e-01
    

    

      C(religion, Treatment('christian'))[T.atheist_spiritual]
       3.739104e-02
        1.495481e-28
       8.054083e-01
    

    

      C(religion, Treatment('christian'))[T.buddhist]
       3.584632e-05
        5.895757e-04
       4.838615e-01
    

    

      C(religion, Treatment('christian'))[T.hindu]
       1.860232e-03
        5.825954e-03
       8.893516e-01
    

    

      C(religion, Treatment('christian'))[T.jewish]
       4.271339e-24
        6.341905e-59
       2.388538e-01
    

    

      C(religion, Treatment('christian'))[T.mixed]
       5.188181e-01
        3.075779e-25
       6.151503e-01
    

    

      C(religion, Treatment('christian'))[T.muslim]
       7.175871e-04
        5.560101e-02
       3.297670e-04
    

    

      C(religion, Treatment('christian'))[T.other]
       1.289143e-01
        2.720419e-09
       1.508855e-09
    

  




















In [83]:



    


result.bse



    


















    
Out[83]:










  
    

      
      0
      1
      2
    

  
  
    

      Intercept
       0.006202
       0.011463
       0.023241
    

    

      C(gender, Treatment('Male'))[T.Female]
       0.018364
       0.025041
       0.052995
    

    

      C(nationality, Treatment('USA'))[T.african_american]
       0.130506
       0.175269
       4.835439
    

    

      C(nationality, Treatment('USA'))[T.asian]
       0.094768
       0.110084
       0.383090
    

    

      C(nationality, Treatment('USA'))[T.hispanic_latino]
       0.086752
       0.133199
       0.411630
    

    

      C(nationality, Treatment('USA'))[T.indigenous]
       0.424156
       0.629396
       1.072631
    

    

      C(nationality, Treatment('USA'))[T.middle eastern]
       0.116193
       0.139561
       0.506906
    

    

      C(nationality, Treatment('USA'))[T.mixed]
       0.218475
       0.319840
       1.008821
    

    

      C(nationality, Treatment('USA'))[T.white]
       0.031707
       0.045642
       0.106449
    

    

      C(religion, Treatment('christian'))[T.asian religion]
       0.451190
       0.658296
       5.171662
    

    

      C(religion, Treatment('christian'))[T.atheist_spiritual]
       0.130460
       0.122884
       0.414588
    

    

      C(religion, Treatment('christian'))[T.buddhist]
       0.354062
       0.240159
       1.006336
    

    

      C(religion, Treatment('christian'))[T.hindu]
       0.211053
       0.281585
       1.012939
    

    

      C(religion, Treatment('christian'))[T.jewish]
       0.067554
       0.081876
       0.357879
    

    

      C(religion, Treatment('christian'))[T.mixed]
       0.228910
       0.183098
       1.008068
    

    

      C(religion, Treatment('christian'))[T.muslim]
       0.191289
       0.218228
       0.331621
    

    

      C(religion, Treatment('christian'))[T.other]
       0.178663
       0.202101
       0.297795

Objectives

  1. Change base case--make traditional, male, white and christian base cases
  2. Covariate plot matrix (religion, gender, nationality)--percentages and maybe correlations
  3. Redo regression analyses (review how to interpret info)--try mixed and accounting for multiple identities
  4. Stacked area plot for all demographics (religion, nationality, gender)

Content source: aboSamoor/compsocial

Similar notebooks:

notebook.community | gallery | about