如何切片 Pandas 日期时间索引以始终选择第一天

Question

If I have a dataframe that looks like this:

timestamp            lat        lon             sog                                     
2018-12-22 08:03:15  54.917200  13.358867      10.8                     
2018-12-22 08:04:05  54.916350  13.354700      10.8                     
2018-12-22 08:04:53  54.915533  13.350767      10.7  
2018-12-24 08:03:15  57.917200  9.358867       10.8                     
2018-12-24 08:04:05  57.916350  9.354700       10.8                     
2018-12-26 08:04:53  59.915533  5.350767       10.7                     

I would like to make a slice of the df to only work with the information from the first day. If I know the date I can simply

first_day = df.loc['2018-12-22']

But I would like to automatically do this for a lot of .csv files and they all have different starting dates. Is there a way to slice the Datetime index to get the first day only? something like:

first_day = df.iloc[:day1] ?

and end up with:

timestamp            lat        lon             sog                                     
2018-12-22 08:03:15  54.917200  13.358867      10.8                     
2018-12-22 08:04:05  54.916350  13.354700      10.8                     
2018-12-22 08:04:53  54.915533  13.350767      10.7  

Answer 1

You can find the first day (min date) and use boolean indexing:

for df in df_list:
    days = df.index.normalize()
    min_date = days.min()
    df[days == min_date]