给定一些k线选股指标，如何挑选符合条件的公司，以python + windquant为例？
【申明：本例只用来作为python学习交流之用，切勿以此作为投资的选股条件】

0、用以下条件挑选公司：

仅作示例用：
【1】本周1 相对于上周五：close上跳，跳空
【2】周2，周3两天，跳空都没有补起来【没有回调——close周2，close周3 > high上周五】
【3】close周2， close周3 > ma20
【4】close周2 > close周1 or close周3 > close周1

1、连接服务器

from WindPy import w
w.start()
w.isconnected()
print(w.isconnected())

2、获取全部A股的股票代码

all_stocks = w.wset("sectorconstituent", "date=2022-05-18;sector=全部A股",usedf=True)
all_stocks_df = all_stocks[1]
all_stocks_df.head(10)

3、在已经过去的日期系列里面，找到离【今天】最近的【周一、周二、周三】日期组，并输出日期。

from datetime import date, timedelta

'''
给定一个指定的日期，从一个日期系列里面，找到此前最近的[周一，周二，周三]组成的系列，输出日期系列
测试代码：
today = date(2023,5,17)
monday, tuesday, wednesday = find_weekday_dates(today)
print(monday, tuesday, wednesday)
'''
def find_weekday_dates(today):
    # 计算今天是星期几（0代表星期一，6代表星期日）
    today_weekday = today.weekday()
    print(today_weekday)
    if(today_weekday<=4):    #周3
        today_weekday += 7
    

    # 计算距离最近的周一、周二、周三的日期
    monday = today - timedelta(days=today_weekday)
    tuesday = monday + timedelta(days=1)
    wednesday = monday + timedelta(days=2)

    return monday, tuesday, wednesday

测试代码：
给定的【今天】是【2023-05-17】找到最近的【周一，周二，周三】

today = date(2023,5,17)
monday, tuesday, wednesday = find_weekday_dates(today)
print(monday, tuesday, wednesday)

2023-05-08 2023-05-09 2023-05-10

计算上个周五的日期

from datetime import datetime, timedelta

# 计算前3天的日期——上个周五的日期
last_friday = monday - timedelta(days=3)
last_friday

4、提取ohlc数据，计算指标，判断条件

计算起始日期和结束日期(today)

today = date.today()
last_month_today = today - timedelta(days = 40)
last_month_today

提取40个日历日的数据

df = w.wsd("000001.SZ", "close,open,low,high", last_month_today, today, "", usedf=True)[1]
df

5、计算ma20的值

# 计算 MA20 数据
ma20 = df["CLOSE"].rolling(window=20).mean()

# 将 MA20 数据添加到 DataFrame 中
df["MA20"] = ma20
df

6、条件判断

【1】本周1 相对于上周五：close上跳，跳空
【2】周2，周3两天，跳空都没有补起来【没有回调——close周2，close周3 > high上周五】
【3】close周2， close周3 > ma20
【4】close周2 > close周1 or close周3 > close周1

#monday, tuesday, wednesday last_friday
cond1 = df.loc[monday].CLOSE > df.loc[last_friday].HIGH
cond2 = (df.loc[tuesday].CLOSE > df.loc[last_friday].HIGH) and (df.loc[wednesday].CLOSE > df.loc[last_friday].HIGH)
cond3 = (df.loc[tuesday].CLOSE > df.loc[tuesday].MA20) and (df.loc[wednesday].CLOSE > df.loc[wednesday].MA20)
cond4 = (df.loc[tuesday].CLOSE > df.loc[monday].CLOSE) and (df.loc[wednesday].CLOSE > df.loc[monday].CLOSE)
print(cond1,cond2,cond3,cond4)

7、合并功能：【下载k线】【计算指标】【判断条件】

"""
功能：获取k线数据，并计算指标，然后用指标来判断选股条件，并返回计算结果
参数：
    stockID：股票代码
    last_month_today：获取k线时的起始日期
    today：获取k线时的结束日期
返回：[stockID,cond1,cond2,cond3,cond4]
    cond1:本周1 相对于上周五：close上跳，跳空
    cond2:周2，周3两天，跳空都没有补起来【没有回调——close周2，close周3  > high上周五】
    cond3:close周2， close周3 > ma20
    cond4:close周2 > close周1  or close周3 > close周1
案例：
    res = calculate_condition("000001.SZ",last_month_today,today)
    res
"""
def calculate_condition(stockID,last_month_today,today):
    df = w.wsd(stockID, "close,open,low,high", last_month_today, today, "", usedf=True)[1]
    
    # 计算 MA20 数据
    ma20 = df["CLOSE"].rolling(window=20).mean()
    df["MA20"] = ma20
    
    #计算选股条件是否满足
    #monday, tuesday, wednesday last_friday
    cond1 = df.loc[monday].CLOSE > df.loc[last_friday].HIGH
    cond2 = (df.loc[tuesday].CLOSE > df.loc[last_friday].HIGH) and (df.loc[wednesday].CLOSE > df.loc[last_friday].HIGH)
    cond3 = (df.loc[tuesday].CLOSE > df.loc[tuesday].MA20) and (df.loc[wednesday].CLOSE > df.loc[wednesday].MA20)
    cond4 = (df.loc[tuesday].CLOSE > df.loc[monday].CLOSE) and (df.loc[wednesday].CLOSE > df.loc[monday].CLOSE)
    #print(cond1,cond2,cond3,cond4)
    return[stockID,cond1,cond2,cond3,cond4]   

8、计算股池的股票，挑选公司

import pandas as pd
res_df = pd.DataFrame(columns = ['stockID','cond1','cond2','cond3','cond4'])
i= 0
for stockid in all_stock_code.values:
    i+=1
    print(i)
    #计算指标和条件
    rtn = calculate_condition(stockid,last_month_today,today)
    new_row_data = {'stockID': rtn[0], 'cond1': rtn[1], 'cond2': rtn[2], 'cond3': rtn[3], 'cond4': rtn[4]}
    res_df = res_df.append(new_row_data,ignore_index = True)
print("计算结束")