from sklearn.datasets import make_classification创建分类数据集

 ### 创建模型

    
 def create_model():
    
     
    
     # 生成数据
    
     from sklearn.datasets import make_classification
    
     X, y = make_classification(n_samples=10000,        # 样本个数
    
                            n_features=25,          # 特征个数
    
                            n_informative=3,        # 有效特征个数
    
                            n_redundant=2,          # 冗余特征个数（有效特征的随机组合）
    
                            n_repeated=0,           # 重复特征个数（有效特征和冗余特征的随机组合）
    
                            n_classes=3,            # 样本类别
    
                            n_clusters_per_class=1, # 簇的个数
    
                            random_state=0)
    
     
    
     print("原始特征维度",X.shape)
    
     
    
     # 读取数据
    
     print("读取数据")
    
     #import pandas as pd
    
     #data = pd.read_csv(datapath)
    
     
    
     # 数据划分
    
     print("数据划分")
    
     from sklearn.model_selection import train_test_split
    
     global x_train,x_valid,x_test,y_train,y_valid,y_test
    
     x_train,x_test,y_train,y_test = train_test_split(X,y,random_state = 33,test_size = 0.25)
    
     x_train,x_valid,y_train,y_valid = train_test_split(x_train,y_train,random_state = 33,test_size = 0.25)
    
  
    
     # 创建模型
    
     print("创建模型")
    
     from sklearn.linear_model import LogisticRegression
    
     global model 
    
     model = LogisticRegression(penalty = 'l2').fit(x_train,y_train)
    
  
    
 ### 保存模型    
    
 def save_model():
    
     print("保存模型")
    
     from sklearn.externals import joblib
    
     joblib.dump(model,'model.pkl')
    
  
    
 ### 模型验证   
    
 def validate_model():
    
     print("模型验证")
    
     print(model.score(x_valid,y_valid))  
    
     
    
 ### 模型预测
    
 def predict_model():
    
     print("模型预测")
    
     global pred
    
     pred = model.predict_proba(x_test)
    
     print(pred)
    
     
    
 if __name__ == "__main__":
    
     create_model()
    
     save_model()
    
     validate_model()
    
     predict_model()