从timedelta python减去天数

Question

I have a DataFrame that looks like this:

raw_data = {'Series_Date':['2017-03-10','2017-03-13','2017-03-14','2017-03-15'],'SeriesDate':['2017-03-10','2017-03-13','2017-03-14','2017-03-15']}
import pandas as pd
df = pd.DataFrame(raw_data,columns=['Series_Date','SeriesDate'])
df['SeriesDate'] = pd.to_datetime(df['SeriesDate'])
df['StartDate'] = df['SeriesDate'] - timedelta(days=10)
df['EndDate'] = df['SeriesDate'] - timedelta(days=3)
print df

When I try to do this:

df['DateDifference'] = ((df.StartDate- df.EndDate) / 2).dt.ceil('D')

I get the following error:

TypeError: ufunc subtract cannot use operands with types dtype('<M8[ns]') and dtype('O')

Any idea how I can overcome this?

Answer 1

you can do this:

import pandas as pd
from  dateutil import relativedelta
raw_data = {'Series_Date':
                ['2017-03-10','2017-03-13','2017-03-14','2017-03-15'],
            'SeriesDate':['2017-03-10','2017-03-13','2017-03-14','2017-03-15']}

df = pd.DataFrame(raw_data,columns=['Series_Date','SeriesDate'])
df['SeriesDate'] = pd.to_datetime(df['SeriesDate'])

move_days = (-10)
df['StartDate'] = (df['SeriesDate']
                   .apply(lambda x: 
                            x + relativedelta.relativedelta(days=move_days)))

movedays= (-3)
df['EndDate'] = (df['SeriesDate']
                 .apply(lambda x: 
                           x + relativedelta.relativedelta(days=move_days)))
print(df)

gives:

   Series_Date SeriesDate  StartDate    EndDate
0  2017-03-10   2017-03-10 2017-02-28 2017-02-28
1  2017-03-13   2017-03-13 2017-03-03 2017-03-03
2  2017-03-14   2017-03-14 2017-03-04 2017-03-04
3  2017-03-15   2017-03-15 2017-03-05 2017-03-05

to get date diff in days,

df['DateDiff'] = (df['EndDate'] - df['StartDate'])
df['DateDiffDays'] = df['DateDiff'].dt.days
print(df)

gives

   Series_Date SeriesDate  StartDate    EndDate  DateDiff  DateDiffDays
0  2017-03-10  2017-03-10 2017-02-28 2017-03-07    7 days             7
1  2017-03-13  2017-03-13 2017-03-03 2017-03-10    7 days             7
2  2017-03-14  2017-03-14 2017-03-04 2017-03-11    7 days             7
3  2017-03-15  2017-03-15 2017-03-05 2017-03-12    7 days             7