In [28]:

    

import numpy as np
import pandas as pd

%matplotlib inline
import matplotlib.pyplot as plt
import seaborn as sns

pd.options.display.max_columns=1000


    

In [29]:

    

df = pd.read_csv('/gh/data/hcmst/1_cleaned.csv', index_col=0)
df.head(4)


    

    
Out[29]:







  

    

      

      
age
      
education
      
race
      
sex
      
household_size
      
house_type
      
income
      
marital_status
      
in_metro
      
usa_region
      
house_payment
      
N_child
      
work
      
has_internet
      
has_gay_friendsfam
      
politics
      
religion
      
in_relationship
      
partner_age
      
N_minutes_survey
      
is_lgb
      
is_married
      
partner_race
      
partner_religion
      
partner_education
      
USA_raised
      
N_marriages
      
cohabit
      
age_first_met
      
age_relationship_begin
      
age_married
      
relative_income
      
same_high_school
      
same_college
      
same_hometown
      
age_difference
      
relationship_quality
      
met_online
      
met_friends
      
met_family
      
met_work
      
relationship_excellent
      
is_not_working
      
has_gay_friends
      
has_gay_family
      
religion_is_christian
      
religion_is_none
      
partner_religion_is_christian
      
partner_religion_is_none
    
  
  

    

      
0
      
52
      
bachelor's degree or higher
      
hispanic
      
female
      
2
      
apartment
      
22250.0
      
living with partner
      
True
      
midwest
      
rent
      
0
      
working - as a paid employee
      
True
      
yes, friends
      
democrat
      
catholic
      
True
      
48.0
      
20
      
True
      
False
      
white
      
protestant (e.g. methodist, lutheran, presbyte...
      
some college
      
True
      
1
      
False
      
45.0
      
45.0
      
NaN
      
less
      
False
      
False
      
False
      
4.0
      
good
      
True
      
False
      
False
      
False
      
False
      
False
      
True
      
False
      
True
      
False
      
False
      
False
    
    

      
1
      
28
      
bachelor's degree or higher
      
white
      
female
      
2
      
apartment
      
45000.0
      
living with partner
      
True
      
west
      
rent
      
0
      
working - as a paid employee
      
True
      
yes, both
      
democrat
      
jewish
      
True
      
30.0
      
13
      
True
      
True
      
white
      
none
      
bachelor's degree or higher
      
True
      
1
      
False
      
19.0
      
20.0
      
23.0
      
more
      
False
      
True
      
False
      
2.0
      
good
      
False
      
True
      
False
      
False
      
False
      
False
      
True
      
True
      
False
      
False
      
False
      
True
    
    

      
2
      
49
      
high school
      
black
      
female
      
4
      
apartment
      
37250.0
      
never married
      
True
      
south
      
rent
      
1
      
working - as a paid employee
      
True
      
yes, both
      
democrat
      
baptist-any denomination
      
False
      
NaN
      
0
      
False
      
False
      
NaN
      
NaN
      
less than high school
      
False
      
NaN
      
False
      
NaN
      
NaN
      
NaN
      
NaN
      
False
      
False
      
False
      
NaN
      
NaN
      
False
      
False
      
False
      
False
      
False
      
False
      
True
      
True
      
True
      
False
      
False
      
False
    
    

      
3
      
31
      
some college
      
white
      
male
      
1
      
apartment
      
45000.0
      
never married
      
True
      
south
      
owned
      
0
      
working - as a paid employee
      
True
      
yes, both
      
democrat
      
other non-christian, please specify:
      
True
      
40.0
      
9
      
True
      
False
      
white
      
other non-christian, please specify
      
high school
      
True
      
0
      
False
      
23.0
      
23.0
      
NaN
      
more
      
False
      
False
      
False
      
9.0
      
good
      
True
      
True
      
False
      
False
      
False
      
False
      
True
      
True
      
False
      
False
      
False
      
False
    
  

In [30]:

    

df['met_to_marriage_years'] = df['age_married'] - df['age_first_met']
df['relationship_to_marriage_years'] = df['age_married'] - df['age_relationship_begin']
df['years_since_marriage'] = df['age'] - df['age_married']
df['same_race'] = df['race'] == df['partner_race']
df['same_religion'] = df['religion'] == df['partner_religion']
df['same_education'] = df['education'] == df['partner_education']


    

In [31]:

    

df.dropna(subset=['relationship_quality'], inplace=True)
y = df['relationship_excellent'].values
df.drop(['relationship_excellent', 'relationship_quality'], axis=1, inplace=True)


    

In [78]:

    

# Model settings
model_config = {
    'algorithm': 'logit',
    'param_grid': {'penalty': ['l1', 'l2'], 'C': np.logspace(-6,3,10)},
#     'algorithm': 'rf',
#     'param_grid': {'n_estimators': [100, 1000], 'max_depth': [5, 7, 9], 'max_features': [.1, .3, .5]},
    'n_folds': 5
}


    

In [79]:

    

from sklearn.pipeline import make_pipeline, Pipeline, FeatureUnion
from sklearn.preprocessing import StandardScaler, OneHotEncoder, LabelEncoder
from sklearn.base import BaseEstimator, TransformerMixin
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.ensemble import RandomForestClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import f1_score, roc_auc_score
from sklearn.feature_extraction import DictVectorizer
from sklearn.preprocessing import StandardScaler,Imputer

from lime.lime_tabular import LimeTabularExplainer


    

In [80]:

    

class DataFrame_Encoder(BaseEstimator, TransformerMixin):
    
    def __init__(self, categorical_cols_=None,numeric_cols_=None):
        self.categorical_cols_ = categorical_cols_
        self.numeric_cols_ = numeric_cols_
    
    def fit(self, df, y=None):
        ### df should be a dataframe that is a mix of categorical and numeric columns
        self.vec_ = DictVectorizer(sparse=False)
        temp_data = df[self.categorical_cols_].astype(str)
        self.vec_.fit(temp_data.to_dict('records'))
        
        self.feature_names_ = list(self.numeric_cols_) + list(self.vec_.feature_names_)
        return self

    def transform(self, df):
        ### df should be a dataframe that is a mix of categorical and numeric columns
        if len(self.categorical_cols_) > 0:
            temp_data = df[self.categorical_cols_].astype(str)
            categorical_data = self.vec_.transform(temp_data.to_dict('records'))
            categorical_df = pd.DataFrame(categorical_data, columns=self.vec_.feature_names_, index=df.index)

            new_data = pd.concat([df[self.numeric_cols_], categorical_df],axis=1)
        else:
            new_data = df.copy()
        return new_data


    

In [81]:

    

# Train and test sets
X_train, X_test, y_train, y_test = train_test_split(df, y, test_size=.2, random_state=0)

# Algorithm type
if model_config['algorithm'] == 'lgbm':
    base_clf = lightgbm.LGBMClassifier(is_unbalance=True, nthread=4)
if model_config['algorithm'] == 'logit':
    base_clf = LogisticRegression(solver='liblinear')
if model_config['algorithm'] == 'rf':
    base_clf = RandomForestClassifier(n_jobs=-1)

# Pipeline
if model_config['algorithm'] == 'rf':
    pipeline = Pipeline([
            ('feature_encoder',DataFrame_Encoder()),
            ('imputer',Imputer()),
            ('clf',base_clf)
    ])
else:
    pipeline = Pipeline([
        ('feature_encoder',DataFrame_Encoder()),
        ('imputer',Imputer()),
        ('scaler',StandardScaler()),
        ('clf',base_clf)
    ])


# Pipeline parameters
string_features = X_train.select_dtypes(include=[object]).columns.values
numeric_features = X_train.select_dtypes(exclude=[object]).columns.values
pipeline.set_params(feature_encoder__categorical_cols_=string_features,
                    feature_encoder__numeric_cols_=numeric_features)

pipeline_params = {}
for key in model_config['param_grid']:
    pipeline_params['clf__'+key] = model_config['param_grid'][key]


    

In [82]:

    

pipeline


    

    
Out[82]:





Pipeline(steps=[('feature_encoder', DataFrame_Encoder(categorical_cols_=array(['education', 'race', 'sex', 'house_type', 'marital_status',
       'usa_region', 'house_payment', 'work', 'has_gay_friendsfam',
       'politics', 'religion', 'partner_race', 'partner_religion',
       'partner_education', 'N_marr...ty='l2', random_state=None, solver='liblinear', tol=0.0001,
          verbose=0, warm_start=False))])

In [83]:

    

pipeline_params


    

    
Out[83]:





{'clf__C': array([  1.00000000e-06,   1.00000000e-05,   1.00000000e-04,
          1.00000000e-03,   1.00000000e-02,   1.00000000e-01,
          1.00000000e+00,   1.00000000e+01,   1.00000000e+02,
          1.00000000e+03]), 'clf__penalty': ['l1', 'l2']}

In [84]:

    

clf = GridSearchCV(pipeline, pipeline_params, cv=model_config['n_folds'],
                   n_jobs=4, verbose=4, scoring='roc_auc')
clf.fit(X_train, y_train)


    

    




Fitting 5 folds for each of 20 candidates, totalling 100 fits
[CV] clf__C=1e-06, clf__penalty=l1 ...................................
[CV] clf__C=1e-06, clf__penalty=l1 ...................................
[CV] clf__C=1e-06, clf__penalty=l1 ...................................
[CV] clf__C=1e-06, clf__penalty=l1 ...................................
[CV] clf__C=1e-06, clf__penalty=l1 ...................................
[CV] clf__C=1e-06, clf__penalty=l2 ...................................
[CV] clf__C=1e-06, clf__penalty=l2 ...................................
[CV] clf__C=1e-06, clf__penalty=l2 ...................................






    




Process ForkPoolWorker-31:
Process ForkPoolWorker-30:
Process ForkPoolWorker-29:
Process ForkPoolWorker-32:
Traceback (most recent call last):
Traceback (most recent call last):
Traceback (most recent call last):
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 249, in _bootstrap
    self.run()
Traceback (most recent call last):
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 249, in _bootstrap
    self.run()
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 249, in _bootstrap
    self.run()
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 249, in _bootstrap
    self.run()
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/pool.py", line 108, in worker
    task = get()
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/pool.py", line 108, in worker
    task = get()
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/pool.py", line 108, in worker
    task = get()
  File "/Users/scott/anaconda/lib/python3.6/site-packages/sklearn/externals/joblib/pool.py", line 360, in get
    racquire()
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/pool.py", line 108, in worker
    task = get()
  File "/Users/scott/anaconda/lib/python3.6/site-packages/sklearn/externals/joblib/pool.py", line 360, in get
    racquire()
  File "/Users/scott/anaconda/lib/python3.6/site-packages/sklearn/externals/joblib/pool.py", line 360, in get
    racquire()
  File "/Users/scott/anaconda/lib/python3.6/site-packages/sklearn/externals/joblib/pool.py", line 362, in get
    return recv()
KeyboardInterrupt
KeyboardInterrupt
KeyboardInterrupt
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/connection.py", line 250, in recv
    buf = self._recv_bytes()
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/connection.py", line 407, in _recv_bytes
    buf = self._recv(4)
  File "/Users/scott/anaconda/lib/python3.6/multiprocessing/connection.py", line 379, in _recv
    chunk = read(handle, remaining)
KeyboardInterrupt






    




---------------------------------------------------------------------------
KeyboardInterrupt                         Traceback (most recent call last)
<ipython-input-84-7d32a738cde7> in <module>()
      1 clf = GridSearchCV(pipeline, pipeline_params, cv=model_config['n_folds'],
      2                    n_jobs=4, verbose=4, scoring='roc_auc')
----> 3 clf.fit(X_train, y_train)

~/anaconda/lib/python3.6/site-packages/sklearn/model_selection/_search.py in fit(self, X, y, groups)
    943             train/test set.
    944         """
--> 945         return self._fit(X, y, groups, ParameterGrid(self.param_grid))
    946 
    947 

~/anaconda/lib/python3.6/site-packages/sklearn/model_selection/_search.py in _fit(self, X, y, groups, parameter_iterable)
    562                                   return_times=True, return_parameters=True,
    563                                   error_score=self.error_score)
--> 564           for parameters in parameter_iterable
    565           for train, test in cv_iter)
    566 

~/anaconda/lib/python3.6/site-packages/sklearn/externals/joblib/parallel.py in __call__(self, iterable)
    766                 # consumption.
    767                 self._iterating = False
--> 768             self.retrieve()
    769             # Make sure that we get a last message telling us we are done
    770             elapsed_time = time.time() - self._start_time

~/anaconda/lib/python3.6/site-packages/sklearn/externals/joblib/parallel.py in retrieve(self)
    717                     ensure_ready = self._managed_backend
    718                     backend.abort_everything(ensure_ready=ensure_ready)
--> 719                 raise exception
    720 
    721     def __call__(self, iterable):

~/anaconda/lib/python3.6/site-packages/sklearn/externals/joblib/parallel.py in retrieve(self)
    680                 # check if timeout supported in backend future implementation
    681                 if 'timeout' in getfullargspec(job.get).args:
--> 682                     self._output.extend(job.get(timeout=self.timeout))
    683                 else:
    684                     self._output.extend(job.get())

~/anaconda/lib/python3.6/multiprocessing/pool.py in get(self, timeout)
    600 
    601     def get(self, timeout=None):
--> 602         self.wait(timeout)
    603         if not self.ready():
    604             raise TimeoutError

~/anaconda/lib/python3.6/multiprocessing/pool.py in wait(self, timeout)
    597 
    598     def wait(self, timeout=None):
--> 599         self._event.wait(timeout)
    600 
    601     def get(self, timeout=None):

~/anaconda/lib/python3.6/threading.py in wait(self, timeout)
    549             signaled = self._flag
    550             if not signaled:
--> 551                 signaled = self._cond.wait(timeout)
    552             return signaled
    553 

~/anaconda/lib/python3.6/threading.py in wait(self, timeout)
    293         try:    # restore state no matter what (e.g., KeyboardInterrupt)
    294             if timeout is None:
--> 295                 waiter.acquire()
    296                 gotit = True
    297             else:

KeyboardInterrupt: 

In [75]:

    

y_train_pred = [x[1] for x in clf.predict_proba(X_train)]
y_test_pred = [x[1] for x in clf.predict_proba(X_test)]
print('AUC=', roc_auc_score(y_test, y_test_pred))


    

    




AUC= 0.635638782577

In [ ]:

    

frac_flag = .1


    

In [68]:

    

df_y = pd.DataFrame({'label': y_test, 'pred': y_test_pred})
sns.violinplot(y='pred', x='label', data=df_y)


    

    




/Users/scott/anaconda/lib/python3.6/site-packages/seaborn/categorical.py:588: FutureWarning: remove_na is deprecated and is a private function. Do not use.
  kde_data = remove_na(group_data)
/Users/scott/anaconda/lib/python3.6/site-packages/seaborn/categorical.py:816: FutureWarning: remove_na is deprecated and is a private function. Do not use.
  violin_data = remove_na(group_data)






    
Out[68]:





<matplotlib.axes._subplots.AxesSubplot at 0x11fc84898>






    

In [ ]: