# __author__ = 'heyin'
# __date__ = '2018/11/16 11:15'
import pandas as pd
from sklearn.feature_extraction import DictVectorizer
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.ensemble import RandomForestClassifier
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import classification_report


def randomf():
    # 获取数据
    t = pd.read_csv('./titanic.csv')
    # 对age列存在的nan进行均值的填充
    t.fillna(t['age'].mean(), inplace=True)
    # 获取pclass，sex，age列
    # print(t.info())  # 通过info就可以看到哪些字段有nan了

    x = t.loc[:, ['pclass', 'sex', 'age']]  # 特征值
    y = t.loc[:, 'survived']  # 目标值series

    # 特征需要处理，利用one-hot编码
    # 将dataframe转为字典
    x = x.to_dict(orient='records')  # records表示，每行数据变成一个字典
    # 划分训练集和测试集
    x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25, random_state=1)
    # 字典的特征抽取
    dv = DictVectorizer(sparse=False)
    x_train = dv.fit_transform(x_train)
    x_test = dv.transform(x_test)
    # print(x_train)
    # print(dv.get_feature_names())
    # 通过交叉验证网格搜索进行超参数调优

    # 使用随机森林进行处理
    rfc = RandomForestClassifier()
    params = {'n_estimators': [80, 100, 120, 200, 300, 500], 'max_depth': [5, 8, 15, 25, 30]}
    gsc = GridSearchCV(rfc, params, cv=2)
    gsc.fit(x_train, y_train)
    print('测试集效果score', gsc.score(x_test, y_test))
    print('训练集效果score', gsc.score(x_train, y_train))
    print(gsc.best_estimator_)


def r_stock():
    # 获取数据
    # 从csv文件获取数据
    df = pd.read_csv('./stockdata/sh.csv')
    df.pop('date')
    y = df.pop('up_down')
    x = df
    # 特征工程需要拆分训练集和测试集后进行
    x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25, random_state=1)
    # 数据标准化处理
    std = StandardScaler()
    x_train = std.fit_transform(x_train)
    x_test = std.transform(x_test)

    # 使用随机森林进行处理
    rfc = RandomForestClassifier()
    params = {'n_estimators': [80, 100, 120, 200, 300, 500], 'max_depth': [5, 8, 15, 25, 30]}
    gsc = GridSearchCV(rfc, params, cv=5)
    gsc.fit(x_train, y_train)
    print('测试集效果score', gsc.score(x_test, y_test))
    print('训练集效果score', gsc.score(x_train, y_train))
    print(gsc.best_estimator_)


def b_stock():
    # 通过上边得到的最佳参数来看召回率和精确率
    # 获取数据
    # 从csv文件获取数据
    df = pd.read_csv('./stockdata/sh.csv')
    df.pop('date')
    y = df.pop('up_down')
    x = df
    # 特征工程需要拆分训练集和测试集后进行
    x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25, random_state=1)
    # 数据标准化处理
    std = StandardScaler()
    x_train = std.fit_transform(x_train)
    x_test = std.transform(x_test)

    # 使用随机森林进行处理
    rfc = RandomForestClassifier(n_estimators=200, max_depth=5)
    rfc.fit(x_train, y_train)
    y_pred = rfc.predict(x_test)
    print(y_pred)
    print('测试集效果score', rfc.score(x_test, y_test))
    print('训练集效果score', rfc.score(x_train, y_train))
    print(classification_report(y_test, y_pred=rfc.predict(x_test), labels=[0, 1],
                                target_names=['跌', '涨']))


if __name__ == '__main__':
    # randomf()
    b_stock()
    # r_stock()