repeated_SMOTE

repeated_SMOTE.py

oversampler = MulticlassOversampling(sv.TRIM_SMOTE(proportion = 0.1))
warnings.filterwarnings("ignore")
Scores1 = []
cmatrices1 = []
cmatrices2 = []
Scores2 = [] 
for i in range(50):
  print("Trial {}".format(i))
  print("-----------------------------")
  scores1 = [] 
  scores2 = []
  trainx, valx, trainy, valy = train_test_split(X.values, Y.values, test_size = 0.25, 
                                                shuffle = True, 
                                                stratify = Y.values,
                                                random_state = i)
  reg = LogisticRegression()
  reg.fit(trainx, trainy)
  s = matthews_corrcoef(valy, reg.predict(valx))
  Scores1.append(s)
  cmatrices1.append(confusion_matrix(valy, reg.predict(valx)))
  
  trainx2, trainy2 = oversampler.sample(trainx, trainy)
  reg2 = LogisticRegression()
  reg2.fit(trainx2, trainy2)
  s = matthews_corrcoef(valy, reg2.predict(valx))
  Scores2.append(s)
  cmatrices2.append(confusion_matrix(valy, reg2.predict(valx)))