select dataframe rows (datetimeindex) by a list of datetime.date

问题

the df looks like this:

DateTime
2017-07-10 03:00:00    288.0
2017-07-10 04:00:00    306.0
2017-08-10 05:00:00    393.0
2017-08-10 06:00:00    522.0
2017-09-10 07:00:00    487.0
2017-09-10 08:00:00    523.0
2017-10-10 09:00:00    585.0

Question how to select row that in a list of dates:

['2017-07-10', '2017-09-10']

to have:

DateTime
2017-07-10 03:00:00    288.0
2017-07-10 04:00:00    306.0
2017-09-10 07:00:00    487.0
2017-09-10 08:00:00    523.0

Thanks

回答1:

Assuming the Datetime is index, try with the below:

to_search=['2017-07-10', '2017-09-10']
df[df.index.to_series().dt.date.astype(str).isin(to_search)]

---

                        1
DateTime                  
2017-07-10 03:00:00  288.0
2017-07-10 04:00:00  306.0
2017-09-10 07:00:00  487.0
2017-09-10 08:00:00  523.0

回答2:

Given that the dates in your list contain up to the daily information, you could start by flooring (Series.dt.floor) the DatetimeIndex up to the daily level and indexing with the list of datetime objects using isin:

t = [pd.to_datetime('2017-07-10'), pd.to_datetime('2017-09-10')]
df.index= pd.to_datetime(df.index)

df[df.index.floor('d').isin(t)]

Output

   DateTime
2017-07-10 03:00:00     288.0
2017-07-10 04:00:00     306.0
2017-09-10 07:00:00     487.0
2017-09-10 08:00:00     523.0

来源：https://stackoverflow.com/questions/55416387/select-dataframe-rows-datetimeindex-by-a-list-of-datetime-date