5

I have a pandas DataFrame with a DateTimeIndex:

                           A          B
2016-04-25 18:50:06   440.967796   201.049600  
2016-04-25 18:50:13   441.054995   200.767034  
2016-04-25 18:50:20   441.142337   200.484475
...
2016-07-27 18:50:06   440.967796   201.049600  
2016-07-27 18:50:13   441.054995   200.767034  
2016-07-27 18:50:20   441.142337   200.484475

I would like to extract all the data of a given date yyyy-mm-dd using a list of dates: ['2016-04-25','2016-04-28',...]

I tried the following:

 df[df.index.isin(['2016-04-25', '2016-04-26'])]

 Empty DataFrame

I would like to retrieve all the data (data of the whole day) of the dates given in this list

4

1 回答 1

8

您需要先通过此解决方案删除时间：

df = df[df.index.normalize().isin(['2016-04-25', '2016-04-26'])]

df = df[df.index.floor('D').isin(['2016-04-25', '2016-04-26'])]

另一种解决方案是比较DatetimeIndex.date，但必须numpy.in1d使用isin：

df = df[np.in1d(df.index.date, pd.to_datetime(['2016-04-25', '2016-04-26']).date)]

或比较创建的字符串DatetimeIndex.strftime：

df = df[np.in1d(df.index.strftime('%Y-%m-%d'), ['2016-04-25', '2016-04-26'])]

print (df)
                              A           B
2016-04-25 18:50:06  440.967796  201.049600
2016-04-25 18:50:13  441.054995  200.767034
2016-04-25 18:50:20  441.142337  200.484475
于 2017-08-29T17:35:05.823 回答