Dates in timeseries models

In [1]:
from __future__ import print_function
import statsmodels.api as sm
import numpy as np
import pandas as pd

Getting started

In [2]:
data = sm.datasets.sunspots.load()

Right now an annual date series must be datetimes at the end of the year.

In [3]:
from datetime import datetime
dates = sm.tsa.datetools.dates_from_range('1700', length=len(data.endog))

Using Pandas

Make a pandas TimeSeries or DataFrame

In [4]:
endog = pd.Series(data.endog, index=dates)

Instantiate the model

In [5]:
ar_model = sm.tsa.AR(endog, freq='A')
pandas_ar_res = ar_model.fit(maxlag=9, method='mle', disp=-1)

Out-of-sample prediction

In [6]:
pred = pandas_ar_res.predict(start='2005', end='2015')
print(pred)

2005-12-31    20.003298
2006-12-31    24.703996
2007-12-31    20.026133
2008-12-31    23.473641
2009-12-31    30.858566
2010-12-31    61.335414
2011-12-31    87.024635
2012-12-31    91.321196
2013-12-31    79.921585
2014-12-31    60.799495
2015-12-31    40.374852
Freq: A-DEC, dtype: float64

Using explicit dates

In [7]:
ar_model = sm.tsa.AR(data.endog, dates=dates, freq='A')
ar_res = ar_model.fit(maxlag=9, method='mle', disp=-1)
pred = ar_res.predict(start='2005', end='2015')
print(pred)

[20.00329845 24.70399631 20.02613267 23.47364059 30.8585664  61.33541408
 87.02463499 91.32119576 79.92158511 60.79949541 40.37485169]

This just returns a regular array, but since the model has date information attached, you can get the prediction dates in a roundabout way.

In [8]:
print(ar_res.data.predict_dates)

DatetimeIndex(['2005-12-31', '2006-12-31', '2007-12-31', '2008-12-31',
               '2009-12-31', '2010-12-31', '2011-12-31', '2012-12-31',
               '2013-12-31', '2014-12-31', '2015-12-31'],
              dtype='datetime64[ns]', freq='A-DEC')

Note: This attribute only exists if predict has been called. It holds the dates associated with the last call to predict.