notebook.community

Edit and run

Classification





In [1]:



    


import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.metrics import accuracy_score, f1_score, confusion_matrix
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVC

%matplotlib inline

Carregamento das amostras





In [10]:



    


df_dados= pd.read_csv('../data/features.csv', sep=';', index_col=0)
df_dados.head()



    


















    
Out[10]:










  
    

      
      nu_CPFCNPJ
      qtdAditivosPorCPFCNPJ
      qtdContratos
      vl_TotalContrato
      vl_Contrato
      participacoes
      valor_total_pregao
      valor_total_tomada
      valor_total_convite
      participacoes_pregao
      participacoes_tomada
      participacoes_convite
      label
      label_pred
      confianca
    

  
  
    

      0
      10140642000140
      26
      8
      415668.77
      71257.0
      5
      0.0
      0.0
      71257.0
      0.0
      0.0
      5.0
      0
      0
      1.0
    

    

      1
      10142026000128
      19
      3
      67887.20
      64587.2
      2
      0.0
      0.0
      64587.2
      0.0
      0.0
      2.0
      0
      0
      1.0
    

    

      2
      10158142000135
      11
      1
      80900.00
      80900.0
      1
      0.0
      0.0
      80900.0
      0.0
      0.0
      1.0
      0
      0
      1.0
    

    

      3
      10171194000141
      4
      1
      45285.00
      45285.0
      1
      0.0
      0.0
      45285.0
      0.0
      0.0
      1.0
      0
      0
      1.0
    

    

      4
      10175041000172
      45
      4
      144250.00
      133450.0
      5
      31050.0
      0.0
      102400.0
      1.0
      0.0
      4.0
      0
      0
      1.0
    

  




















In [3]:



    


samples = df_dados.drop(['nu_CPFCNPJ', 'label'], axis=1).values
labels = df_dados.label.values
print(samples.shape, labels.shape)



    


















    






(2075, 11) (2075,)

Classificação





In [4]:



    


estimators = [('std', StandardScaler()), ('svm', SVC(probability=True))]
pipe = Pipeline(estimators)
parameters = dict(svm__C=[1,10], svm__kernel=['linear', 'rbf'])

clf = GridSearchCV(pipe, parameters, scoring='f1', verbose=10, cv=10)
clf.fit(samples, labels)
print(clf.best_params_, clf.best_score_)



    


















    






Fitting 10 folds for each of 4 candidates, totalling 40 fits
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.964912, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.955752, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.944444, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................










    






[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.1s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.2s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:    0.3s remaining:    0.0s










    






[CV] ..... svm__kernel=linear, svm__C=1, score=0.920354, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.963636, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.955752, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.907407, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................










    






[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.3s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   6 out of   6 | elapsed:    0.4s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   7 out of   7 | elapsed:    0.4s remaining:    0.0s
[Parallel(n_jobs=1)]: Done   8 out of   8 | elapsed:    0.5s remaining:    0.0s










    






[CV] ..... svm__kernel=linear, svm__C=1, score=0.914286, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.929825, total=   0.1s
[CV] svm__kernel=linear, svm__C=1 ....................................
[CV] ..... svm__kernel=linear, svm__C=1, score=0.880734, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................










    






[Parallel(n_jobs=1)]: Done   9 out of   9 | elapsed:    0.6s remaining:    0.0s










    






[CV] ........ svm__kernel=rbf, svm__C=1, score=0.954955, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.982143, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.963636, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.972477, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.944444, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.972477, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.915888, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.912621, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.972973, total=   0.1s
[CV] svm__kernel=rbf, svm__C=1 .......................................
[CV] ........ svm__kernel=rbf, svm__C=1, score=0.915888, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.964912, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.938053, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.954128, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.939130, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.955752, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.955752, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.907407, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.934579, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.939130, total=   0.1s
[CV] svm__kernel=linear, svm__C=10 ...................................
[CV] .... svm__kernel=linear, svm__C=10, score=0.880734, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.964286, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.982143, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.982143, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.982456, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.982143, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.981818, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.927273, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.953271, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.972973, total=   0.1s
[CV] svm__kernel=rbf, svm__C=10 ......................................
[CV] ....... svm__kernel=rbf, svm__C=10, score=0.945455, total=   0.1s
{'svm__kernel': 'rbf', 'svm__C': 10} 0.967414311082










    






[Parallel(n_jobs=1)]: Done  40 out of  40 | elapsed:    3.6s finished

















In [6]:



    


best_c, best_kernel = 10, 'rbf'



    











In [7]:



    


df_dados['label_pred'] = clf.predict(samples)
df_dados.head()



    


















    
Out[7]:










  
    

      
      nu_CPFCNPJ
      qtdAditivosPorCPFCNPJ
      qtdContratos
      vl_TotalContrato
      vl_Contrato
      participacoes
      valor_total_pregao
      valor_total_tomada
      valor_total_convite
      participacoes_pregao
      participacoes_tomada
      participacoes_convite
      label
      label_pred
    

  
  
    

      0
      10140642000140
      26
      8
      415668.77
      71257.0
      5
      0.0
      0.0
      71257.0
      0.0
      0.0
      5.0
      0
      0
    

    

      1
      10142026000128
      19
      3
      67887.20
      64587.2
      2
      0.0
      0.0
      64587.2
      0.0
      0.0
      2.0
      0
      0
    

    

      2
      10158142000135
      11
      1
      80900.00
      80900.0
      1
      0.0
      0.0
      80900.0
      0.0
      0.0
      1.0
      0
      0
    

    

      3
      10171194000141
      4
      1
      45285.00
      45285.0
      1
      0.0
      0.0
      45285.0
      0.0
      0.0
      1.0
      0
      0
    

    

      4
      10175041000172
      45
      4
      144250.00
      133450.0
      5
      31050.0
      0.0
      102400.0
      1.0
      0.0
      4.0
      0
      0
    

  




















In [8]:



    


probs = clf.predict_proba(samples)
df_dados['confianca'] = np.around(probs[:,0], decimals=2)
df_dados.head()



    


















    
Out[8]:










  
    

      
      nu_CPFCNPJ
      qtdAditivosPorCPFCNPJ
      qtdContratos
      vl_TotalContrato
      vl_Contrato
      participacoes
      valor_total_pregao
      valor_total_tomada
      valor_total_convite
      participacoes_pregao
      participacoes_tomada
      participacoes_convite
      label
      label_pred
      confianca
    

  
  
    

      0
      10140642000140
      26
      8
      415668.77
      71257.0
      5
      0.0
      0.0
      71257.0
      0.0
      0.0
      5.0
      0
      0
      1.0
    

    

      1
      10142026000128
      19
      3
      67887.20
      64587.2
      2
      0.0
      0.0
      64587.2
      0.0
      0.0
      2.0
      0
      0
      1.0
    

    

      2
      10158142000135
      11
      1
      80900.00
      80900.0
      1
      0.0
      0.0
      80900.0
      0.0
      0.0
      1.0
      0
      0
      1.0
    

    

      3
      10171194000141
      4
      1
      45285.00
      45285.0
      1
      0.0
      0.0
      45285.0
      0.0
      0.0
      1.0
      0
      0
      1.0
    

    

      4
      10175041000172
      45
      4
      144250.00
      133450.0
      5
      31050.0
      0.0
      102400.0
      1.0
      0.0
      4.0
      0
      0
      1.0
    

  




















In [9]:



    


df_dados.to_csv(path_or_buf='../data/features.csv', sep=';')



    











In [ ]:

Content source: DiegoPereira/Teupassadotecondena

Similar notebooks:

notebook.community | gallery | about