基于sklearn的机器学习股票预测

ゝ一世哀愁。 2023-01-16 04:57 248阅读 0赞

## main\_linear.py: ##

import pandas as pd
    import numpy as np
    import sys
    import os
    import requests
    from concurrent.futures import ThreadPoolExecutor
    from datetime import date, timedelta, datetime
    
    from pandas import DataFrame
    from sklearn.model_selection import train_test_split
    from sklearn.preprocessing import StandardScaler
    from sklearn.linear_model import LogisticRegression, LinearRegression
    import joblib
    
    
    def download_code_data(*args):
        code, start_date, end_date, data_path, predict_path = args
        code_data_length_limit = 2000
        code_download_path_urls = (
            "http://quotes.money.163.com/service/chddata.html?code=1{code}&start={start}&end={end}&fields={fields}",
            "http://quotes.money.163.com/service/chddata.html?code=0{code}&start={start}&end={end}&fields={fields}"
        )
        fields = 'TCLOSE;HIGH;LOW;TOPEN;LCLOSE;CHG;PCHG;TURNOVER;VOTURNOVER;VATURNOVER;TCAP;MCAP'
        for url_template in code_download_path_urls:
            u = url_template.format(code=code, start=start_date, end=end_date, fields=fields)
            headers = { 
                'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/85.0.4183.83 Safari/537.36',
                'Accept-Language': 'zh-CN,zh;q=0.9',
                'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9',
                'Accept-Encoding': 'gzip, deflate'
            }
            response = requests.get(u, headers=headers).text
            # 转成数组
            if len(response) >= code_data_length_limit:
                text = requests.get(u, headers=headers).text.split("\r\n")
                for i in range(len(text), 3, -1):
                    now = text[i - 1 - 1].split(",")[8]
                    next = text[i - 1 - 1 - 1].split(",")[8]
    
                    if next.lower() == "none" or now.lower() == "none":
                        text[i - 1 - 1] = ""
                        continue
    
                    # 处理日期
                    today = text[i - 1 - 1].split(",\'")[0]
                    yueri = today.split("-")[1] + today.split("-")[2]
                    text[i - 1 - 1] = yueri + text[i - 1 - 1].split(today)[1]
                # 获取今天的数据
                today_data = text[1]
                today_year = today_data.split(",\'")[0]
                today_month = today_year.split("-")[1] + today_year.split("-")[2]
                today_data = today_month + today_data.split(today_year)[1]
                text[0] = ""
                text[1] = ""
                with open(os.path.join(data_path, '{code}.csv'.format(code=code)), 'w', encoding='utf-8') as f:
                    file_str = ""
                    for data in text:
                        file_str = file_str + data + "\r\n"
                    f.write(file_str)
                print("股票编号：" + code + " 数据爬取成功!\n")
                train_and_predict(code, today_data)
                return True
        return False
    
    
    # path = r'C:\Users\Dell\Desktop\cal\cml\3\xinan\xinan'
    # p = read_csv(path, sep='\s+', encoding='ANSI', header=None)
    # c = pd.concat(p) # 将p里面全部dataframe合成一个
    
    
    def train_and_predict(code, today_data):
        # 读取文件
        column_names = ['日期', '股票代码', '名称', '收盘价', '最高价', '最低价', '开盘价', '前收盘',
                        '涨跌额', '涨跌幅', '换手率', '成交量', '成交金额', '总市值', '流通市值']
    
        data = pd.read_csv(end_date + "\\data\\" + code + ".csv", names=column_names)
        # 数据处理，去除？转换为nan
        data.replace(to_replace="?", value=np.nan)
        # 删除nan的数据
        data.dropna(inplace=True)
        # 取特征值、目标值
        x = data.iloc[:, [0, 3, 4, 5, 6, 7, 8, 10, 11, 12, 13, 14]]
        y = data["涨跌幅"]
        # 划分训练集
        x_train, x_test, y_train, y_test = train_test_split(x, y)
        # 特征工程，标准化
        transfer = StandardScaler()
        x_train = transfer.fit_transform(x_train)
        x_test = transfer.transform(x_test)
        if os.path.exists("model.pkl"):
            esetimator = joblib.load("model1.pkl")
        else:
            esetimator = LinearRegression()
        esetimator.fit(x_train, y_train)
        y_predict = esetimator.predict(x_test)
    
        # print("y_predict:", y_predict)
        # print("直接比对：\n", y_test, y_predict)
        score = esetimator.score(x_test, y_test)
        # print("精准率：\n", score)
        # 保存模型
        joblib.dump(esetimator, "model1.pkl")
        # 预测明天的情况
        x_today = DataFrame([today_data.split(",")])
        x_today = x_today.iloc[:, [0, 3, 4, 5, 6, 7, 8, 10, 11, 12, 13, 14]]
        x_today = transfer.transform(x_today)
        t_predict = esetimator.predict(x_today)
        if t_predict[0] > 0:
            result = "股票编号：" + code + "\r\n股票名称：" + today_data.split(",")[2] + "\r\n下个交易日会涨" + str(t_predict) + "，精确率：" + str(
                score)
        else:
            result = "股票编号：" + code + "\r\n股票名称：" + today_data.split(",")[2] + "\r\n下个交易日会跌" + str(t_predict) + "，精确率：" + str(
                score)
        print(result)
        if score > 0.7:
            with open(os.path.join(end_date + "\\predict_linear", code + '.csv'.format(code=code)), 'w',
                      encoding='utf-8') as f:
                f.write(result)
    
    
    if __name__ == '__main__':
        if len(sys.argv) > 1:
            all_code = sys.argv[1:]
            # print(all_code)
        else:
            all_code_url = "http://44.push2.eastmoney.com/api/qt/clist/get?pn=1&pz=10000&po=1&np=1&ut=bd1d9ddb04089700cf9c27f6f7426281&fltt=2&invt=2&fid=f3&fs=m:0+t:6,m:0+t:13,m:0+t:80,m:1+t:2,m:1+t:23&fields=f12&_=1579615221139"
            r = requests.get(all_code_url, timeout=5).json()
            all_code = [data['f12'] for data in r['data']['diff']]
            print(all_code)
        start_date = (date.today() - timedelta(days=720)).strftime("%Y%m%d")
        end_date = date.today().strftime("%Y%m%d")
        data_path = os.path.join(end_date, "data")
        os.makedirs(data_path, exist_ok=True)
        predict_path = os.path.join(end_date, "predict_linear")
        os.makedirs(predict_path, exist_ok=True)
        # print(start_date, end_date)
        # 下载数据
        # with ThreadPoolExecutor(max_workers=20) as tpe:
        # tpe.map(download_code_data, [(code, start_date, end_date, data_path, predict_path) for code in all_code])
        # start()
        # for code in all_code:
        download_code_data("688777", start_date, end_date, data_path, predict_path)

## main\_logistic.py: ##

import pandas as pd
    import numpy as np
    import sys
    import os
    import requests
    from concurrent.futures import ThreadPoolExecutor
    from datetime import date, timedelta, datetime
    
    from pandas import DataFrame
    from sklearn.model_selection import train_test_split
    from sklearn.preprocessing import StandardScaler
    from sklearn.linear_model import LogisticRegression, LinearRegression
    import joblib
    
    
    def download_code_data(*args):
        code, start_date, end_date, data_path, predict_path = args
        code_data_length_limit = 20000
        code_download_path_urls = (
            "http://quotes.money.163.com/service/chddata.html?code=1{code}&start={start}&end={end}&fields={fields}",
            "http://quotes.money.163.com/service/chddata.html?code=0{code}&start={start}&end={end}&fields={fields}"
        )
        fields = 'TCLOSE;HIGH;LOW;TOPEN;LCLOSE;CHG;PCHG;TURNOVER;VOTURNOVER;VATURNOVER;TCAP;MCAP'
        for url_template in code_download_path_urls:
            u = url_template.format(code=code, start=start_date, end=end_date, fields=fields)
            headers = { 
                'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/85.0.4183.83 Safari/537.36',
                'Accept-Language': 'zh-CN,zh;q=0.9',
                'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9',
                'Accept-Encoding': 'gzip, deflate'
            }
            response = requests.get(u, headers=headers).text
            # 转成数组
            if len(response) >= code_data_length_limit:
                text = requests.get(u, headers=headers).text.split("\r\n")
                for i in range(len(text), 3, -1):
                    now = text[i - 1 - 1].split(",")[8]
                    next = text[i - 1 - 1 - 1].split(",")[8]
    
                    if next.lower() == "none" or now.lower() == "none":
                        continue
                    if float(next) > 0:
                        text[i - 1 - 1] = text[i - 1 - 1] + "," + "1"
                    else:
                        text[i - 1 - 1] = text[i - 1 - 1] + "," + "0"
    
                    # 处理日期
                    today = text[i - 1 - 1].split(",\'")[0]
                    yueri = today.split("-")[1] + today.split("-")[2]
                    text[i - 1 - 1] = yueri + text[i - 1 - 1].split(today)[1]
                # 获取今天的数据
                today_data = text[1]
                today_year = today_data.split(",\'")[0]
                today_month = today_year.split("-")[1] + today_year.split("-")[2]
                today_data = today_month + today_data.split(today_year)[1]
                text[0] = ""
                text[1] = ""
                with open(os.path.join(data_path, '{code}.csv'.format(code=code)), 'w', encoding='utf-8') as f:
                    file_str = ""
                    for data in text:
                        file_str = file_str + data + "\r\n"
                    f.write(file_str)
                print("股票编号：" + code + " 数据爬取成功!\n")
                train_and_predict(code, today_data)
                return True
        return False
    
    
    # path = r'C:\Users\Dell\Desktop\cal\cml\3\xinan\xinan'
    # p = read_csv(path, sep='\s+', encoding='ANSI', header=None)
    # c = pd.concat(p) # 将p里面全部dataframe合成一个
    
    
    def train_and_predict(code, today_data):
        # 读取文件
        column_names = ['日期', '股票代码', '名称', '收盘价', '最高价', '最低价', '开盘价', '前收盘',
                        '涨跌额', '涨跌幅', '换手率', '成交量', '成交金额', '总市值', '流通市值', 'Class']
    
        data = pd.read_csv(end_date + "\\data\\" + code + ".csv", names=column_names)
        # 数据处理，去除？转换为nan
        data.replace(to_replace="?", value=np.nan)
        # 删除nan的数据
        data.dropna(inplace=True)
        # 取特征值、目标值
        x = data.iloc[:, [0, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]]
        y = data["Class"]
        # 划分训练集
        x_train, x_test, y_train, y_test = train_test_split(x, y)
        # 特征工程，标准化
        transfer = StandardScaler()
        x_train = transfer.fit_transform(x_train)
        x_test = transfer.transform(x_test)
        if os.path.exists("model.pkl"):
            esetimator = joblib.load("model.pkl")
        else:
            esetimator = LogisticRegression()
        esetimator.fit(x_train, y_train)
        y_predict = esetimator.predict(x_test)
    
        print("y_predict:", y_predict)
        print("直接比对：\n", y_test == y_predict)
        score = esetimator.score(x_test, y_test)
        print("精准率：\n", score)
        # 保存模型
        joblib.dump(esetimator, "model1.pkl")
        # 预测明天的情况
        x_today = DataFrame([today_data.split(",")])
        x_today = x_today.iloc[:, [0, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]]
        x_today = transfer.transform(x_today)
        t_predict = esetimator.predict(x_today)
        if t_predict[0] > 0:
            result = "股票编号：" + code + "\r\n股票名称：" + today_data.split(",")[2] + "\r\n下个交易日会涨" + str(t_predict) + "，精确率：" + str(
                score)
        else:
            result = "股票编号：" + code + "\r\n股票名称：" + today_data.split(",")[2] + "\r\n下个交易日会跌" + str(t_predict) + "，精确率：" + str(
                score)
        print(result)
        if score > 0.7:
            with open(os.path.join(end_date + "\\predict_logistic", code + '.csv'.format(code=code)), 'w',
                      encoding='utf-8') as f:
                f.write(result)
    
    
    if __name__ == '__main__':
        if len(sys.argv) > 1:
            all_code = sys.argv[1:]
            print(all_code)
        else:
            all_code_url = "http://44.push2.eastmoney.com/api/qt/clist/get?pn=1&pz=10000&po=1&np=1&ut=bd1d9ddb04089700cf9c27f6f7426281&fltt=2&invt=2&fid=f3&fs=m:0+t:6,m:0+t:13,m:0+t:80,m:1+t:2,m:1+t:23&fields=f12&_=1579615221139"
            r = requests.get(all_code_url, timeout=5).json()
            all_code = [data['f12'] for data in r['data']['diff']]
            print(all_code)
        start_date = (date.today() - timedelta(days=720)).strftime("%Y%m%d")
        end_date = date.today().strftime("%Y%m%d")
        data_path = os.path.join(end_date, "data")
        os.makedirs(data_path, exist_ok=True)
        predict_path = os.path.join(end_date, "predict_logistic")
        os.makedirs(predict_path, exist_ok=True)
        print(start_date, end_date)
        # 下载数据
        # with ThreadPoolExecutor(max_workers=20) as tpe:
        # tpe.map(download_code_data, [(code, start_date, end_date, data_path, predict_path) for code in all_code])
        # start()
        for code in all_code:
            download_code_data(code, start_date, end_date, data_path, predict_path)