In [1]:

    

import pydotplus 
import pandas as pd

from IPython.display import Image
from sklearn import tree, grid_search
from sklearn.cross_validation import train_test_split, StratifiedShuffleSplit
from sklearn.model_selection import cross_val_score
from sklearn.metrics import accuracy_score, classification_report, roc_auc_score
import numpy as np
import matplotlib.pyplot as plt


    

    




/home/daniel/anaconda3/lib/python3.5/site-packages/sklearn/cross_validation.py:44: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.
  "This module will be removed in 0.20.", DeprecationWarning)
/home/daniel/anaconda3/lib/python3.5/site-packages/sklearn/grid_search.py:43: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. This module will be removed in 0.20.
  DeprecationWarning)

In [2]:

    

# %pylab inline


    

In [3]:

    

german = pd.read_csv('german_credit.csv', header=0, sep=',')


    

In [4]:

    

german.head()


    

    
Out[4]:






  

    

      

      
Creditability
      
Account Balance
      
Duration of Credit (month)
      
Payment Status of Previous Credit
      
Purpose
      
Credit Amount
      
Value Savings/Stocks
      
Length of current employment
      
Instalment per cent
      
Sex and Marital Status
      
...
      
Duration in Current address
      
Most valuable available asset
      
Age (years)
      
Concurrent Credits
      
Type of apartment
      
No of Credits at this Bank
      
Occupation
      
No of dependents
      
Telephone
      
Foreign Worker
    
  
  

    

      
0
      
1
      
1
      
18
      
4
      
2
      
1049
      
1
      
2
      
4
      
2
      
...
      
4
      
2
      
21
      
3
      
1
      
1
      
3
      
1
      
1
      
1
    
    

      
1
      
1
      
1
      
9
      
4
      
0
      
2799
      
1
      
3
      
2
      
3
      
...
      
2
      
1
      
36
      
3
      
1
      
2
      
3
      
2
      
1
      
1
    
    

      
2
      
1
      
2
      
12
      
2
      
9
      
841
      
2
      
4
      
2
      
2
      
...
      
4
      
1
      
23
      
3
      
1
      
1
      
2
      
1
      
1
      
1
    
    

      
3
      
1
      
1
      
12
      
4
      
0
      
2122
      
1
      
3
      
3
      
3
      
...
      
2
      
1
      
39
      
3
      
1
      
2
      
2
      
2
      
1
      
2
    
    

      
4
      
1
      
1
      
12
      
4
      
0
      
2171
      
1
      
3
      
4
      
3
      
...
      
4
      
2
      
38
      
1
      
2
      
2
      
2
      
1
      
1
      
2
    
  

5 rows × 21 columns

In [5]:

    

german_target = german.Creditability.values

german_data = german.iloc[:, 1:]


    

In [6]:

    

sum(german_target) / float(len(german_target))


    

    
Out[6]:





0.69999999999999996

In [7]:

    

german_data.columns


    

    
Out[7]:





Index(['Account Balance', 'Duration of Credit (month)',
       'Payment Status of Previous Credit', 'Purpose', 'Credit Amount',
       'Value Savings/Stocks', 'Length of current employment',
       'Instalment per cent', 'Sex and Marital Status', 'Guarantors',
       'Duration in Current address', 'Most valuable available asset',
       'Age (years)', 'Concurrent Credits', 'Type of apartment',
       'No of Credits at this Bank', 'Occupation', 'No of dependents',
       'Telephone', 'Foreign Worker'],
      dtype='object')

In [8]:

    

model = tree.DecisionTreeClassifier(max_depth=2)
model.fit(german_data, german_target)
dot_data = tree.export_graphviz(model, out_file="tree3.out", 
                         feature_names=german_data.columns,  
                         class_names=['credit', 'no credit'],  
                         filled=True, rounded=True,  
                         special_characters=False)  
graph = pydotplus.graphviz.graph_from_dot_file("tree3.out")  
Image(graph.create_png())


    

    
Out[8]:

In [9]:

    

model = tree.DecisionTreeClassifier()
model.fit(german_data, german_target)

dot_data = tree.export_graphviz(model, out_file="big_tree.out", 
                         feature_names=german_data.columns,  
                         class_names=['0', '1'],  
                         filled=True, rounded=True,  
                         special_characters=True)  

graph = pydotplus.graphviz.graph_from_dot_file("big_tree.out")
Image(graph.create_png())


    

    
Out[9]:

In [10]:

    

train_data, test_data, train_target, test_target = train_test_split(german_data, german_target,
                                                                   test_size = 0.2)


    

In [11]:

    

model.fit(train_data, train_target)

test_predictions = model.predict(test_data)

print(roc_auc_score(test_target, test_predictions))

train_predictions = model.predict(train_data)

print(roc_auc_score(train_target, train_predictions))


    

    




0.684573652553
1.0

In [12]:

    

print(classification_report(test_target, test_predictions))


    

    




             precision    recall  f1-score   support

          0       0.53      0.61      0.56        61
          1       0.82      0.76      0.79       139

avg / total       0.73      0.71      0.72       200

In [13]:

    

np.mean(cross_val_score(model, german_data, german_target, cv=5))


    

    
Out[13]:





0.67299999999999993

In [14]:

    

def f_cv(i):
    model = tree.DecisionTreeClassifier(max_depth=i)
    return np.mean(cross_val_score(model, german_data, german_target, cv=5, scoring='roc_auc'))

def f_test(i):
    model = tree.DecisionTreeClassifier(max_depth=i)
    model.fit(train_data, train_target)
    test_predictions = model.predict(test_data)
    return roc_auc_score(test_target, test_predictions)

def f_train(i):
    model = tree.DecisionTreeClassifier(max_depth=i)
    model.fit(train_data, train_target)
    return roc_auc_score(test_target, test_predictions)

plt.figure(figsize=(10, 10))
plt.title('score(depth)')

plt.grid(True)
plt.ylim((0.0, 1.0))
plt.plot([i for i in range(1, 31)], [i for i in map(f_cv, range(1, 31))], label='cross val score')
plt.plot([i for i in range(1, 31)], [i for i in map(f_test, range(1, 31))], label='test')
plt.xlabel('depth')
plt.ylabel('score')
plt.legend(loc='best')
plt.show()

# pylab.legend(loc='lower right')


    

In [ ]: