自定义 Pandas 日历使其可迭代

huangapple
go评论100阅读模式
英文:

Custom calendar in Pandas make iterable

问题

在下面的 NYSE 自定义日历中，使用 pandas 试图使该类可迭代，以运行以下检查：

  1. nyse_cal = NYSECalendar()
  2. for date in nyse_cal:
  3.     print(date)

但语句 last_holiday = self.rules[-1] 由于索引 [-1] 而导致 TypeError: 'method' is not subscriptable，这是因为针对 'Holiday' 实例进行了索引。是否有解决方法？

  1. from pandas.tseries.offsets import CustomBusinessDay, MonthEnd
  2. from pandas.tseries.holiday import AbstractHolidayCalendar, Holiday, \
  3.     USMemorialDay,  USMartinLutherKingJr, USPresidentsDay, GoodFriday, \
  4.     USLaborDay, USThanksgivingDay, nearest_workday, next_monday
  6. class NYSECalendar(AbstractHolidayCalendar):
  7.     ''' 纽约证券交易所节假日日历 '''
  8.     rules = [
  9.         Holiday('New Years Day', month=1, day=1, observance=nearest_workday),
  10.         USMartinLutherKingJr,
  11.         USPresidentsDay,
  12.         GoodFriday,
  13.         USMemorialDay,
  14.         Holiday('Juneteenth', month=6, day=19, observance=nearest_workday),
  15.         Holiday('USIndependenceDay', month=7, day=4, observance=nearest_workday),
  16.         USLaborDay,
  17.         USThanksgivingDay,
  18.         Holiday('Christmas', month=12, day=25, observance=nearest_workday),
  19.         ]
  20.     def __iter__(self):
  21.         last_holiday = self.rules[-1]
  22.         last_holiday_date = last_holiday.dates[-1]
  23.         start_date = last_holiday_date + MonthEnd(12) + CustomBusinessDay(calendar=self)
  24.         end_date = start_date + MonthEnd(1) - CustomBusinessDay(calendar=self)
  25.         curr_date = start_date
  26.         while curr_date <= end_date:
  27.             yield curr_date
  28.             curr_date += CustomBusinessDay(calendar=self)
英文:

trying to make the class iterable for the below NYSE custom calendar using pandas, say to run this check:

  1. nyse_cal = NYSECalendar()
  2. for date in nyse_cal:
  3.     print(date)

but the statement last_holiday = self.rules[-1] produces TypeError: &#39;method&#39; is not subscriptable for the 'Holiday' instance because of indexing [-1]. Is there a workaround?

  1. from pandas.tseries.offsets import CustomBusinessDay, MonthEnd
  2. from pandas.tseries.holiday import AbstractHolidayCalendar, Holiday, \
  3.     USMemorialDay,  USMartinLutherKingJr, USPresidentsDay, GoodFriday, \
  4.     USLaborDay, USThanksgivingDay, nearest_workday, next_monday
  6. class NYSECalendar(AbstractHolidayCalendar):
  7.     &#39;&#39;&#39; NYSE holiday calendar via pandas &#39;&#39;&#39;
  8.     rules = [
  9.         Holiday(&#39;New Years Day&#39;, month=1, day=1, observance=nearest_workday),
  10.         USMartinLutherKingJr,
  11.         USPresidentsDay,
  12.         GoodFriday,
  13.         USMemorialDay,
  14.         Holiday(&#39;Juneteenth&#39;, month=6, day=19, observance=nearest_workday),
  15.         Holiday(&#39;USIndependenceDay&#39;, month=7, day=4, observance=nearest_workday),
  16.         USLaborDay,
  17.         USThanksgivingDay,
  18.         Holiday(&#39;Christmas&#39;, month=12, day=25, observance=nearest_workday),
  19.         ]
  20.     def __iter__(self):
  21.         last_holiday = self.rules[-1]
  22.         last_holiday_date = last_holiday.dates[-1]
  23.         start_date = last_holiday_date + MonthEnd(12) + CustomBusinessDay(calendar=self)
  24.         end_date = start_date + MonthEnd(1) - CustomBusinessDay(calendar=self)
  25.         curr_date = start_date
  26.         while curr_date &lt;= end_date:
  27.             yield curr_date
  28.             curr_date += CustomBusinessDay(calendar=self)

答案1

得分: 1

你可以这样做：

  1. from pandas.tseries.holiday import AbstractHolidayCalendar, Holiday, nearest_workday, USMartinLutherKingJr, USPresidentsDay, GoodFriday, USMemorialDay, USLaborDay, USThanksgivingDay
  2. from pandas.tseries.offsets import MonthEnd, CustomBusinessDay
  3. import pandas as pd
  5. class NYSECalendar(AbstractHolidayCalendar):
  6.     rules = [
  7.         Holiday('New Years Day', month=1, day=1, observance=nearest_workday),
  8.         USMartinLutherKingJr,
  9.         USPresidentsDay,
  10.         GoodFriday,
  11.         USMemorialDay,
  12.         Holiday('Juneteenth', month=6, day=19, observance=nearest_workday),
  13.         Holiday('USIndependenceDay', month=7, day=4, observance=nearest_workday),
  14.         USLaborDay,
  15.         USThanksgivingDay,
  16.         Holiday('Christmas', month=12, day=25, observance=nearest_workday),
  17.     ]
  19.     def get_rules(self):
  20.         return self.rules
  22.     def __iter__(self):
  23.         last_holiday = self.get_rules()[-1]
  24.         last_holiday_date = last_holiday.dates[-1]
  25.         start_date = last_holiday_date + MonthEnd(12) + CustomBusinessDay(calendar=self)
  26.         end_date = start_date + MonthEnd(1) - CustomBusinessDay(calendar=self)
  27.         curr_date = start_date
  28.         while curr_date <= end_date:
  29.             yield curr_date
  30.             curr_date += CustomBusinessDay(calendar=self)

并在这个样本数据框中测试它，如下：

  1. import pandas as pd
  3. nyse_cal = NYSECalendar()
  5. start_date = '2022-01-01'
  6. end_date = '2022-01-31'
  8. df = pd.DataFrame({
  9.     'business_day': pd.date_range(start=start_date, end=end_date, freq=CustomBusinessDay(calendar=nyse_cal))
  10. })
  12. print(df)

返回工作日：

  1.   business_day
  2. 0    2022-01-03
  3. 1    2022-01-04
  4. 2    2022-01-05
  5. 3    2022-01-06
  6. 4    2022-01-07
  7. 5    2022-01-10
  8. 6    2022-01-11
  9. 7    2022-01-12
  10. 8    2022-01-13
  11. 9    2022-01-14
  12. 10    2022-01-18
  13. 11    2022-01-19
  14. 12    2022-01-20
  15. 13    2022-01-21
  16. 14    2022-01-24
  17. 15    2022-01-25
  18. 16    2022-01-26
  19. 17    2022-01-27
  20. 18    2022-01-28
  21. 19    2022-01-31

基本上，你需要获取正确的包。

英文:

You cn do this:

  1. from pandas.tseries.holiday import AbstractHolidayCalendar, Holiday, nearest_workday,USMartinLutherKingJr, USPresidentsDay, GoodFriday, USMemorialDay, USLaborDay,USThanksgivingDay
  2. from pandas.tseries.offsets import MonthEnd, CustomBusinessDay
  3. import pandas as pd
  6. class NYSECalendar(AbstractHolidayCalendar):
  7.     rules = [
  8.         Holiday(&#39;New Years Day&#39;, month=1, day=1, observance=nearest_workday),
  9.         USMartinLutherKingJr,
  10.         USPresidentsDay,
  11.         GoodFriday,
  12.         USMemorialDay,
  13.         Holiday(&#39;Juneteenth&#39;, month=6, day=19, observance=nearest_workday),
  14.         Holiday(&#39;USIndependenceDay&#39;, month=7, day=4, observance=nearest_workday),
  15.         USLaborDay,
  16.         USThanksgivingDay,
  17.         Holiday(&#39;Christmas&#39;, month=12, day=25, observance=nearest_workday),
  18.     ]
  19.     
  20.     def get_rules(self):
  21.         return self.rules
  23.     def __iter__(self):
  24.         last_holiday = self.get_rules()[-1]
  25.         last_holiday_date = last_holiday.dates[-1]
  26.         start_date = last_holiday_date + MonthEnd(12) + CustomBusinessDay(calendar=self)
  27.         end_date = start_date + MonthEnd(1) - CustomBusinessDay(calendar=self)
  28.         curr_date = start_date
  29.         while curr_date &lt;= end_date:
  30.             yield curr_date
  31.             curr_date += CustomBusinessDay(calendar=self)

and test it it this sample dataframe, this way:

  1. import pandas as pd
  3. nyse_cal = NYSECalendar()
  5. start_date = &#39;2022-01-01&#39;
  6. end_date = &#39;2022-01-31&#39;
  8. df = pd.DataFrame({
  9.     &#39;business_day&#39;: pd.date_range(start=start_date, end=end_date, freq=CustomBusinessDay(calendar=nyse_cal))
  10. })
  12. print(df)

return th business days:

  1.   business_day
  2. 0    2022-01-03
  3. 1    2022-01-04
  4. 2    2022-01-05
  5. 3    2022-01-06
  6. 4    2022-01-07
  7. 5    2022-01-10
  8. 6    2022-01-11
  9. 7    2022-01-12
  10. 8    2022-01-13
  11. 9    2022-01-14
  12. 10   2022-01-18
  13. 11   2022-01-19
  14. 12   2022-01-20
  15. 13   2022-01-21
  16. 14   2022-01-24
  17. 15   2022-01-25
  18. 16   2022-01-26
  19. 17   2022-01-27
  20. 18   2022-01-28
  21. 19   2022-01-31

Basically, you need to get th right packages.

huangapple
  • 本文由 发表于 2023年2月23日 22:13:58
  • 转载请务必保留本文链接：https://go.coder-hub.com/75545992.html
  • pandas
  • python
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定