提前一天设置熊猫日期时间索引

Question

Is it possible to set one day forward a datetime index in pandas?

I've got this index

mydata.data.index
Out[7]: <class 'pandas.tseries.index.DatetimeIndex'>
[2013-01-27 22:00:00, ..., 2014-12-16 22:00:00]
Length: 500, Freq: None, Timezone: None

and its wrong. The correct date that the data begin is the day after. That is 2014-12-17 and they end at 2013-01-28

Is there a way to fix this?

Answer 1

Some different options:

Use the shift method:

In [10]: dti = pd.date_range('2012-01-01', periods=2)

In [11]: dti
Out[11]:
<class 'pandas.tseries.index.DatetimeIndex'>
[2012-01-01, 2012-01-02]
Length: 2, Freq: D, Timezone: None

In [13]: dti.shift(1, freq='D')
Out[13]:
<class 'pandas.tseries.index.DatetimeIndex'>
[2012-01-02, 2012-01-03]
Length: 2, Freq: D, Timezone: None

But maybe more easy (as @Evert notices), you can simply add a timedelta of one day:

In [14]: dti + pd.Timedelta('1 day')
Out[14]:
<class 'pandas.tseries.index.DatetimeIndex'>
[2012-01-02, 2012-01-03]
Length: 2, Freq: None, Timezone: None

For older pandas versions, you can also just use datetime.timedelta(days=1) ( pd.Timedelta is only avaible starting from 0.15).