将日期列分组为 n 天的时间段
我需要一个函数，根据某些开始日期和结束日期（1 年间隔）将日期列分组为 n 天周期。为了给数据框中的每个日期分配一个季度（~90 天的时间段），我使用了下面的代码，这不是很简洁（而且我也想在 30 天的时间段内重复使用它）
def get_quarter(row, start_date, col_name):
# date = row['TRN_DT']
date = row[col_name]
if date >= start_date and date <= start_date + timedelta(days = 90):
return 0
if date > start_date + timedelta(days = 90) and date <= start_date + timedelta(180):
return 1
if date > start_date + timedelta(180) and date <= start_date + timedelta(270):
return 2
return 3
它基本上逐行检查当前日期所属的间隔。我想知道是否有更好的方法来做到这一点。 pandas.Series.dt.to_period() 不会这样做，因为它使用日历年作为参考 --start 01.Jan, end 31.Dec;也就是说，16.Jan.XXXX 将始终位于 Q1 中；如果开始日期是 6 月 16 日，我想要的是 16.Jan 位于第三季度。谢谢
I need a function that groups date column into n-day periods with respect to some start and end dates (1 year interval). To assign a quarter (~90 day period) to every date in the data frame I used the code below, which is not very neat (and I want to reuse it for 30-day period as well)
def get_quarter(row, start_date, col_name):
# date = row['TRN_DT']
date = row[col_name]
if date >= start_date and date <= start_date + timedelta(days = 90):
return 0
if date > start_date + timedelta(days = 90) and date <= start_date + timedelta(180):
return 1
if date > start_date + timedelta(180) and date <= start_date + timedelta(270):
return 2
return 3
It basically checks row by row which interval current date belongs to. I was wondering whether there is a better way to do this. pandas.Series.dt.to_period() will not do since it uses a calendar year as a reference --start 01.Jan, end 31.Dec; that is, 16.Jan.XXXX will always be in Q1; what I want is for 16.Jan to be in Q3 if the start date is 16-Jun. Thanks
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
FTR，一个可能的解决方案是根据
start_date移动系列中的每个日期，以模拟start_date是年初：我们计算当前日期之间的差值日期和同年年初。
我们将所有日期移至同一个报价以计算“新季度”
FTR, a possible solution is to shift every date in the series according to the
start_date, to simulate thatstart_dateis the beginning of the year:We calculate the difference between the current date and the beginning of the same year.
We move all of our dates by the same offser to calculate the "new quarter"