foreword

ChatGPT is popular all over the Internet, so is there a way to realize AI stock selection without writing formulas? The answer is yes, today we will try Xiaoda of Tongdaxin, and let Xiaoda choose stocks according to our requirements.

1. What is Tongdaxin-Xiaoda?

Tongdaxin-Xiaoda is an investment system that is easy to operate, and can realize the needs of users such as intelligent stock selection, information retrieval, and answering questions by inputting search sentences, making investment a one-sentence matter.
Through artificial intelligence technology, Xiaoda integrates market quotations, information, databases and major functional components into an intelligent service
product with multiple functions including semantic analysis, information aggregation and machine learning, and supports the service docking of its own content.

Tongdaxin-Xiaoda functions include: smart stock selection, stock Haishen search, full-text search, etc.
Official website address: https://wenda.tdx.com.cn/

2. Use steps

This is mainly to use requests, which should all come with the system, and I modified it into a class method, which will be more convenient to call.
In fact, someone has already written the code before, and I also refer to it to modify it. It’s just that you need to update the cookie every time, otherwise you won’t be able to access it, and it’s not very friendly to quant friends, so here’s a little bit, and you can use it conveniently in the future.

1. Introduce browser_cookie3 library

browser_cookie3 can read cookies already cached by computer browsers (such as Chrome Firefox Opera Edge Chromium Brave). Install it before use to facilitate subsequent operations. Of course, you can do it manually, or you don’t need this library. However, for the convenience of use, it is recommended to install it. After all, it is troublesome to log in and extract cookies manually every day.

pip install browser-cookie3 

2. Tongdaxin-Xiaoda AI stock selection source code

The first is to complete the registration, please do it yourself. This article supports the following two access methods at the same time. Method 2 is recommended, because automatic access can be completed as long as the computer logs in with a browser.

Method 1: Use any browser to log in to https://wenda.tdx.com.cn/, and press F12 to view the network, find the cookie, copy it, replace the ASPSessionID value here, and then use it.
xd = TDX_xiaoda(cookie = 'ASPSessionID=xxxxxxxxxxx')

Method 2: After logging in to https://wenda.tdx.com.cn/ with the Chrome browser, there is no need to manually find the cookie, the program will automatically obtain the ASPSessionID and enable access. But you must use the Chrome browser.
xd = TDX_xiaoda()

# -*- coding: utf-8 -*-
'''
通达信-小达
'''
import requests
import browser_cookie3
import pandas as pd
import json,os,random


class TDX_xiaoda:
    '''
    通达信-小达，使用chrome浏览器登录后，此工具可以自动获取ASPSessionID，并实现访问。
    '''
    def __init__(self,cookie=''):
        # 如果类传送Cookie则使用传参
        if cookie=='': 
            self.cookie = self.get_cookie_tdx()
        else:
            self.cookie = cookie

    def set_headers(self):
        # 增加UserAgent，防止被反爬虫拦截
        user_agent_list = ["Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/68.0.3440.106 Safari/537.36",
                            "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36",
                            "Mozilla/5.0 (Windows NT 10.0; WOW64) Gecko/20100101 Firefox/61.0",
                            "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/64.0.3282.186 Safari/537.36",
                            "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/62.0.3202.62 Safari/537.36",
                            "Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/45.0.2454.101 Safari/537.36",
                            "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0)",
                            "Mozilla/5.0 (Macintosh; U; PPC Mac OS X 10.5; en-US; rv:1.9.2.15) Gecko/20110303 Firefox/3.6.15",
                            ]
        headers = {
            'Cookie': self.cookie,
            'User-Agent': random.choice(user_agent_list)}
        return headers

    def _get_cookie_tdx(self):
        # 查找指定域名的指定Cookie值
        cookies = browser_cookie3.chrome(domain_name='wenda.tdx.com.cn')
        for item in cookies:
            if item.name == "ASPSessionID" :
                # print('%s = %s' % (item.name, item.value))
                cookie_tdx = 'ASPSessionID=' + item.value + ';'
                return cookie_tdx

    def get_data_option(self,word='涨停股票'):
        '''
        获取提示参考
        :param word:
        :return:
        '''
        data = [{"op_flag":1,"question":word,"POS":0,"COUNT":10,"RANG":"AG"}]
        url = 'https://wenda.tdx.com.cn/TQL?Entry=NLPSE.QuestionImagine&RI='
        headers = self.set_headers()
        res = requests.post(url=url,data=json.dumps(data),headers=headers)
        res_json = res.json()
        return res_json
        
    def get_word_code(self,name='换手率2.8~15,量比2~20,OBV指标创新高,涨幅2~7,股价2~30'):
        '''
        :param name:
        :return:
        '''
        data = [{"message":name,"TDXID":"","wdbk":"","RANG":"AG"}]
        url = 'https://wenda.tdx.com.cn/TQL?Entry=NLPSE.StockSelect&RI='
        headers = self.set_headers()
        res = requests.post(url=url,headers=headers,data=json.dumps(data))
        res_json = res.json()
        code = res_json[-1][0]
        return code
        
    def get_all_option_data(self):
        '''
        获取全部参考
        :return:
        '''
        url = 'https://wenda.tdx.com.cn/TQL?Entry=NLPSE.SmartQuery&RI='
        headers = self.set_headers()
        data = [{"op_flag":1,"order_field":"","order_flag":1,"cond_json":"","POS":0,"COUNT":-1,"RANG":"AG"}]
        res = requests.post(url=url,headers=headers,data=json.dumps(data))
        res_json = res.json()
        df = pd.DataFrame(res_json)
        df = df.iloc[1:]
        df2 = df.rename(columns=df.iloc[0])
        df3 = df2.iloc[1:]
        return df3
        
    def get_word_result(self,word='今日涨停',try_times = 30):
        '''
        根据关键字分析后获取数据
        :param word:关键字
        :return:字典，包含状态和数据
        '''
        while try_times>0:
            url = 'https://wenda.tdx.com.cn/TQL?Entry=NLPSE.NLPQuery&RI=6BFD'
            code = self.get_word_code(name=word)
            headers = self.set_headers()
            data = [{"nlpse_id":code,"POS":0,"COUNT":100000,"order_field":"","dynamic_order":"","order_flag":"","timestamps":0,"op_flag":1,"screen_type":1,"RANG":"AG"}]
            res = requests.post(url=url,headers=headers,data=json.dumps(data))
            text = res.content.decode('utf-8')
            res_json = json.loads(text)
            if len(res_json[1])<4:
                try_times -= 1
                print('获取信息失败，剩余'+str(try_times)+'次尝试机会，再次尝试连接......')
            else:
                # print(text)
                df = pd.DataFrame(res_json)
                df = df.iloc[1:]
                df2 = df.rename(columns=df.iloc[0])
                df3 = df2.iloc[2:]
                return {
                        "success": True,
                        "df_data":df3.round(3)
                    }
        return {
            "success": False,
            "msg": '获取行情信息结果为None'
        }


# ===============表格美化输出===============
def df_table(df,index):
    import prettytable as pt
    #利用prettytable对输出结果进行美化,index为索引列名:df_table(df,'market')
    tb = pt.PrettyTable()
    # 如果为trade_time为index转换为日期类型，其它不用管。
    if index == "trade_time":
        df = df.set_index(index)
        df.index = pd.DatetimeIndex(df.index)
    # df.reset_index(level=None, drop=True, inplace=True, col_level=0, col_fill='')
    df = df.reset_index(drop = True)
    tb.add_column(index,df.index)#按date排序
    for col in df.columns.values:#df.columns.values的意思是获取列的名称
        # print('col',col)
        # print('df[col]',df[col])
        tb.add_column(col, df[col])
    print(tb)


if __name__=='__main__':
    # 以下两种方法均可以，推荐方法2。注册请自行搞定。
    # 方法1：使用任意浏览器登录https://wenda.tdx.com.cn/，并按F12查看网络，找到Cookie复制后替换这里ASPSessionID后数值即可使用。注意不要带分号。
    # xd = TDX_xiaoda(cookie = 'ASPSessionID=xxxxxxxxxx') 
    # 方法2：使用Chrome浏览器登录https://wenda.tdx.com.cn/后，无需手工查找Cookie，程序会自动获取ASPSessionID，并实现访问。但必须使用Chrome浏览器。
    xd = TDX_xiaoda() 
        
    if 0:
        #可能获取不到
        ret = xd.get_data_option(word='沪深A股,换手率1%~5%')
        print(ret)
        #获取问题转代码
        ret = xd.get_word_code(name='沪深A股,换手率1%~5%')
        print(ret)
        #获取全部的参考，具体看通达信
        ret = xd.get_all_option_data()
        # print(ret)
        df_table(ret,'df')
    
    if 1:
        #获取问题的结果
        words = '沪深A股,周线MACD持续向上,今日涨停' # 连板 3连板 今日涨停 20230201涨停 北向资金 板块 大单抢筹 MACD背离
        ret = xd.get_word_result(word=words)
        if ret['success']:
            df = ret['df_data']
            # print(df)
            df_table(df,'df')
            filename = os.path.dirname(__file__)+'\\'+'通达信-小达选股_'+words+'.xlsx'
            df.to_excel(filename)
            print('选股完成，结果写入：',filename)
            

Summarize

This time I wrote about Tongdaxin’s AI stock selection. I wrote about Wencai AI stock selection using Flush before. The following is the portal. Those who need it can learn by themselves.

[AI stock selection] How to call the wencai package through python to realize AI stock selection (even beginners can learn it) https://blog.csdn.net/popboy29/article/details/126336227

It is also very convenient to ask for money in Flush, but each search is limited to 200 items. It depends on the individual to call. Tongdaxin has not found any restrictions so far. In addition, this does not require installation, the code is transparent, and it is convenient to call.

After working for a day, my shoulders are a little sore, and writing is not easy. The key is to constantly debug the code, which is very troublesome. If you find it helpful, you may wish to give it a like and start debugging. I wish you all a fortune!

At the same time, I would like to express my gratitude to Tongdaxin and the prawn who wrote the code before!