Pandas本周营业日股价数据

Question

我的问题与以下不同：

问题1：Week of a month pandas 问题2：Week number of the month

上述问题涉及假设一周7天。它试图计算每周有7天的时间。我的数据是由（工作日）每日价格组成的，有时会因为假期市场关闭而错过一周中的几天

我的问题是，如果给定一个日期，如何查找一个月中的一周。注：我突出显示了“给定日期”，因为此过程每天都在处理，因此任何看起来在月底之前的答案都可能不起作用

我的尝试是前瞻性的，但不是最佳的：

def is_third_friday(s):
    d = datetime.datetime.strptime(s, '%Y-%m-%d')
    return d.weekday() == 5 and 15 <= d.day <= 21

dow = deepcopy(data['Close'] * np.nan).to_frame()
dow.columns = ['OpexFriday']
dow['next_date'] = pd.Series([str(i.date() + datetime.timedelta(days=1)) for i in dow.index]).values
dow['OpexFriday'] = pd.Series([is_third_friday(str(i)) for i in dow['next_date']]).values
dow['OpexWeek'] = (dow['OpexFriday'] * 1).replace(0, np.nan).fillna(method='bfill', limit=4).replace(np.nan, 0) == True

我不知道如何提供一些示例data，但如果您转到“https://aroussi.com/post/python-yahoo-finance“翻页并使用authors yfinance软件包，您将能够获得一些价格数据

上述函数将查找当月的第三周（均为真）。此外，它还将设定该周的星期五

如果你发现这个问题有任何问题或者是重复的，请告诉我。我已经寻找了一段时间的解决办法

Answer 1

一种方法是使用timedelta将任何日期更改为下一个星期五，然后检查下一个星期五是否在15到21之间

from datetime import datetime, timedelta
def OpexWeek (s):
    d = datetime.strptime(s, '%Y-%m-%d')
    day = (d+timedelta(days=(4-d.weekday())%7)).day
    return (day>=15) & (day<=21)

然后你得到

#for the example the second Friday of June 2020:
OpexWeek('2020-06-12')
False

# the Monday after is True because part of the OpexWeek
OpexWeek('2020-06-15')
True

注意：需要知道的一件事是，OpexWeek之前的周六和周日是真实的，但是因为您说您的数据是工作日，所以这不重要

datetime系列上使用的pandas版本可以是：

def OpexWeekPd (ser):
    return (ser+pd.to_timedelta((4-ser.dt.weekday)%7, unit='d')).dt.day.between(15,21)

举个小例子：

print (
    pd.DataFrame({'date':pd.bdate_range('2020-06-01', '2020-06-30').astype(str)})
      .assign(isOpexWeek=lambda x: x['date'].apply(OpexWeek), 
              isIpexWeekPd=lambda x: OpexWeekPd(pd.to_datetime(x['date'])))
    )
          date  isOpexWeek  isIpexWeekPd
0   2020-06-01       False         False
1   2020-06-02       False         False
2   2020-06-03       False         False
3   2020-06-04       False         False
4   2020-06-05       False         False
5   2020-06-08       False         False
6   2020-06-09       False         False
7   2020-06-10       False         False
8   2020-06-11       False         False
9   2020-06-12       False         False
10  2020-06-15        True          True
11  2020-06-16        True          True
12  2020-06-17        True          True
13  2020-06-18        True          True
14  2020-06-19        True          True
15  2020-06-22       False         False
16  2020-06-23       False         False
17  2020-06-24       False         False
18  2020-06-25       False         False
19  2020-06-26       False         False
20  2020-06-29       False         False
21  2020-06-30       False         False

Answer 2

我们可以轻松地修改您的函数以处理索引：

# sample data
dow = pd.DataFrame(index=pd.date_range('2020-01-01', '2020-01-31'),
                   columns=['OpexFriday'])

isFriday = dow.index.dayofweek == 5
thirdWeek = dow.index.day.to_series().between(15,21)

# third Friday
dow['OpexFriday'] = (isFriday & thirdWeek).values

# third work week
dow['OpexWeek'] = dow['OpexFriday'].where(dow['OpexFriday']).bfill(limit=4).fillna(0)

# extract the third week:
dow[dow['OpexWeek']==1]

输出：

            OpexFriday  OpexWeek
2020-01-14       False       1.0
2020-01-15       False       1.0
2020-01-16       False       1.0
2020-01-17       False       1.0
2020-01-18        True       1.0

Answer 3

import datetime
from math import ceil


def week_of_month(dt):
    """ Returns the week of the month for the specified date.
    """

    adjusted_dom = dt.day + dt.replace(day=1).day

    return int(ceil(adjusted_dom / 7.0))


def week_of_month_from_str(d_str):
    return week_of_month(datetime.datetime.strptime(d_str, '%Y-%m-%d'))


assert week_of_month_from_str("2020-03-02") == 1
assert week_of_month_from_str("2020-03-07") == 2
assert week_of_month_from_str("2020-03-13") == 2
assert week_of_month_from_str("2020-03-14") == 3
assert week_of_month_from_str("2020-03-20") == 3
assert week_of_month_from_str("2020-06-01") == 1
assert week_of_month_from_str("2020-06-06") == 1
assert week_of_month_from_str("2020-06-07") == 2
assert week_of_month_from_str("2020-06-08") == 2