5

I am trying to run a Augmented Dickey-Fuller test in statsmodels in Python, but I seem to be missing something.

This is the code that I am trying:

import numpy as np
import statsmodels.tsa.stattools as ts

x = np.array([1,2,3,4,3,4,2,3])
result = ts.adfuller(x)

I get the following error:

Traceback (most recent call last):
  File "C:\Users\Akavall\Desktop\Python\Stats_models\stats_models_test.py", line 12, in <module>
    result = ts.adfuller(x)
  File "C:\Python27\lib\site-packages\statsmodels-0.4.1-py2.7-win32.egg\statsmodels\tsa\stattools.py", line 201, in adfuller
    xdall = lagmat(xdiff[:,None], maxlag, trim='both', original='in')
  File "C:\Python27\lib\site-packages\statsmodels-0.4.1-py2.7-win32.egg\statsmodels\tsa\tsatools.py", line 305, in lagmat
    raise ValueError("maxlag should be < nobs")
ValueError: maxlag should be < nobs

My Numpy Version: 1.6.1 My statsmodels Version: 0.4.1 I am using windows.

I am looking at the documentation here but can't figure what I am doing wrong. What am I missing?

Thanks in Advance.

bmu
  • 35,119
  • 13
  • 91
  • 108
Akavall
  • 82,592
  • 51
  • 207
  • 251

1 Answers1

7

I figured it out. By default maxlag is set to None, while it should be set to integer. Something like this works:

import numpy as np
import statsmodels.tsa.stattools as ts

x = np.array([1,2,3,4,3,4,2,3])
result = ts.adfuller(x, 1) # maxlag is now set to 1

Output:

>>> result
(-2.6825663173365015, 0.077103947319183241, 0, 7, {'5%': -3.4775828571428571, '1%': -4.9386902332361515, '10%': -2.8438679591836733}, 15.971188911270618)
Akavall
  • 82,592
  • 51
  • 207
  • 251
  • 7
    The problem is that if maxlag is None, then adfuller uses a default maxlag recommended in the literature, int(np.ceil(12. * np.power(nobs/100., 1/4.))), which does not work for small numbers. Similar problem with minimum number of observations in grangercausalitytests https://github.com/statsmodels/statsmodels/issues/347 . – Josef Jun 30 '12 at 16:13