Dates in timeseries models

In [1]:
from __future__ import print_function
import statsmodels.api as sm
import numpy as np
import pandas as pd
/build/statsmodels-IFPJo1/statsmodels-0.8.0/.pybuild/cpython3_3.7_statsmodels/build/statsmodels/compat/pandas.py:56: FutureWarning: The pandas.core.datetools module is deprecated and will be removed in a future version. Please use the pandas.tseries module instead.
  from pandas.core import datetools

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    20.003289
2006-12-31    24.703980
2007-12-31    20.026123
2008-12-31    23.473651
2009-12-31    30.858570
2010-12-31    61.335438
2011-12-31    87.024666
2012-12-31    91.321229
2013-12-31    79.921617
2014-12-31    60.799527
2015-12-31    40.374891
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)
[20.00328884 24.70398005 20.02612319 23.47365066 30.85857032 61.33543769
 87.02466633 91.32122934 79.92161688 60.7995268  40.37489106]

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.