我有两个日期列'StartDate'和'EndDate'。我想查找从2019年12月起的两个日期之间每个月的天数,并向前忽略2019年之前的任何几个月进行计算。每行的StartDate和EndDate可以跨2年,月份重叠,并且Date列也可以为空。
样本数据:
df = {'Id': ['1','2','3','4','5','6','7', '8'],
'Item': ['A','B','C','D','E','F','G', 'H'],
'StartDate': ['2019-12-10', '2019-12-01', '2019-10-01', '2020-01-01', '2019-03-01','2019-03-01','2019-10-01', ''],
'EndDate': ['2020-02-21' ,'2020-01-01','2020-08-31','2020-01-30','2019-12-31','2019-12-31','2020-08-31', '']
}
df = pd.DataFrame(df,columns= ['Id', 'Item','StartDate','EndDate'])
预期O / P:
下面的解决方案部分起作用。
df['StartDate'] = pd.to_datetime(df['StartDate'])
df['EndDate'] = pd.to_datetime(df['EndDate'])
def days_of_month(x):
s = pd.date_range(*x, freq='D').to_series()
return s.resample('M').count().rename(lambda x: x.month)
df1 = df[['StartDate', 'EndDate']].apply(days_of_month, axis=1).fillna(0)
df_final = df[['StartDate', 'EndDate']].join([df['StartDate'].dt.year.rename('Year'), df1])
尝试一下:
df.join(df.dropna(axis=0,how='any')
.apply(lambda x: pd.date_range(x['StartDate'],x['EndDate'], freq='D')
.to_frame().resample('M').count().loc['2019-12-01':].unstack(), axis=1)[0].fillna(0))
输出:
Id Item StartDate EndDate 2019-12-31 00:00:00 2020-01-31 00:00:00 2020-02-29 00:00:00 2020-03-31 00:00:00 2020-04-30 00:00:00 2020-05-31 00:00:00 2020-06-30 00:00:00 2020-07-31 00:00:00 2020-08-31 00:00:00
0 1 A 2019-12-10 2020-02-21 22.0 31.0 21.0 0.0 0.0 0.0 0.0 0.0 0.0
1 2 B 2019-12-01 2020-01-01 31.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
2 3 C 2019-10-01 2020-08-31 31.0 31.0 29.0 31.0 30.0 31.0 30.0 31.0 31.0
3 4 D 2020-01-01 2020-01-30 0.0 30.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4 5 E 2019-03-01 2019-12-31 31.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5 6 F 2019-03-01 2019-12-31 31.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
6 7 G 2019-10-01 2020-08-31 31.0 31.0 29.0 31.0 30.0 31.0 30.0 31.0 31.0
7 8 H NaT NaT NaN NaN NaN NaN NaN NaN NaN NaN NaN
我们将创建两个大的DataFrame,一个在每个月的开始,另一个在每个月的结束。然后,我们将它们相应地裁剪,这给我们提供了一个简单的减法。由于您要包含结束日期,因此我们需要添加1天,然后我们清除所有负日期(应为0)。
import pandas as pd df_s = pd.DataFrame([pd.date_range('2019-12-01', '2020-12-01', freq='MS').to_numpy()], index=df.index) df_e = (pd.DataFrame([pd.date_range('2019-12-01', '2020-12-01', freq='MS').to_numpy()], index=df.index) + pd.offsets.MonthEnd(1)) df_s = df_s.clip(lower=pd.to_datetime(df.StartDate), axis=0) df_e = df_e.clip(upper=pd.to_datetime(df.EndDate), axis=0) res = ((df_e - df_s) + pd.to_timedelta(1, 'd')).clip(lower=pd.to_timedelta(0, 'd')) res.columns = pd.period_range(start='2019-12', end='2020-12', freq='M') # So int for col in res.columns: res[col] = res[col].dt.days df = pd.concat([df, res], axis=1)
Id Item StartDate EndDate 2019-12 2020-01 2020-02 2020-03 2020-04 2020-05 2020-06 2020-07 2020-08 2020-09 2020-10 2020-11 2020-12 0 1 A 2019-12-10 2020-02-21 22.0 31.0 21.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1 2 B 2019-12-01 2020-01-31 31.0 31.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2 3 C 2019-10-01 2020-08-31 31.0 31.0 29.0 31.0 30.0 31.0 30.0 31.0 31.0 0.0 0.0 0.0 0.0 3 4 D 2020-01-01 2020-01-30 0.0 30.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4 5 E 2019-03-01 2019-12-31 31.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5 6 F 2019-03-01 2019-12-31 31.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6 7 G 2019-10-01 2020-08-31 31.0 31.0 29.0 31.0 30.0 31.0 30.0 31.0 31.0 0.0 0.0 0.0 0.0 7 8 H NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN