[ How to combine a hierarchical date and time index into a single datetime index in Pandas? ]

Suppose I have the following dataframe:

                     temp
2014-11-20 12:45:00        51
2014-11-20 12:46:00        43
2014-11-20 12:47:00        44
2014-11-21 12:45:00        44
2014-11-21 12:46:00        46
2014-11-21 12:47:00        48
2014-11-22 12:45:00        38
2014-11-22 12:46:00        32
2014-11-22 12:47:00        37

If you copy the above and use pd.read_clipboard() I believe you should get a hierarchical index (or MultiIndex)

The first index is the date and the second index is the time as strings. How can I combine the two indexes into a single datetime index?

(Note this is not about using pd.read_csv() parse_dates option because the dataframe is not coming from a csv file, the question is how to do it with an existing dataframe.)

Answer 1

You can use the to_datetime method to parse strings to datetime values. To use this, I think the easiest is to first just concatenate the date and time strings leading to a single level string index:

In [184]: s
Out[184]:
                     temp
2014-11-20 12:45:00    51
           12:46:00    43
           12:47:00    44
2014-11-21 12:45:00    44
           12:46:00    46
           12:47:00    48
2014-11-22 12:45:00    38
           12:46:00    32
           12:47:00    37

In [187]: s.index = s.index.get_level_values(0).values + ' ' + s.index.get_level_values(1)

In [188]: s.index
Out[188]: Index([u'2014-11-20 12:45:00', u'2014-11-20 12:46:00', ... u'2014-11-22 12:47:00'], dtype='object')

And then you can use to_datetime to convert the strings to a DatetimeIndex:

In [189]: s.index = pd.to_datetime(s.index)

In [190]: s.index
Out[190]:
<class 'pandas.tseries.index.DatetimeIndex'>
[2014-11-20 12:45:00, ..., 2014-11-22 12:47:00]
Length: 9, Freq: None, Timezone: None