def Elastic_selection(k, X, y):
elasticnet = ElasticNet()
elasticresult = RFE(elasticnet,
k).fit(X, y.values.ravel()).get_support(indices=True)
elasticresult = X[X.columns[elasticresult]]
elasticresult.to_csv("ElasticNet_out.csv")
return None
def L1SVM_selection(k, X, y):
lsvc = LinearSVC(penalty="l1", dual=False)
lsvcresult = RFE(lsvc, k).fit(X,
y.values.ravel()).get_support(indices=True)
lsvcresult = X[X.columns[lsvcresult]]
lsvcresult.to_csv("L1SVM_out.csv")
return None
def Lasso_selection(k, X, y):
clf = LassoCV(cv=k)
lassoresult = RFE(clf, k).fit(X,
y.values.ravel()).get_support(indices=True)
lassoresult = X[X.columns[lassoresult]]
lassoresult.to_csv("Lasso_out.csv")
return None
def ExtraTree_selection(k, X, y):
extraT = ExtraTreesClassifier(n_estimators=10)
extraresult = RFE(extraT,
k).fit(X, y.values.ravel()).get_support(indices=True)
extraresult = X[X.columns[extraresult]]
extraresult.to_csv("ExtraTree_out.csv")
return None
def LOG_RFE_selection(k, X, y):
log = LogisticRegression(solver='liblinear')
logresult = RFE(log, k).fit(X, y.values.ravel()).get_support(indices=True)
logresult = X[X.columns[logresult]]
logresult.to_csv("LOG-FRE_out.csv")
return None
def SVM_RFE_selection(k, X, y):
svc = SVC(kernel="linear")
svmresult = RFE(svc, k).fit(X, y.values.ravel()).get_support(indices=True)
svmresult = X[X.columns[svmresult]]
svmresult.to_csv("SVM-RFE_out.csv")
return None
def XGB_selection(k, X, y):
xgb = XGBClassifier()
xgbresult = RFE(xgb, k).fit(X, y.values.ravel()).get_support(indices=True)
xgbresult = X[X.columns[xgbresult]]
xgbresult.to_csv("XGBoost_out.csv")
return None
