您如何在statsmodels ADF测试中设置要测试的滞后次数？

Question

I know you can specify maxlags in the parameters, but whenever I do the output shows that I am using zero lags. Here's a generic example of the problem I'm having. I set maxlag = 1 in this example. The result however shows that the number of lags used is 0, corresponding to the third parameter in the output.

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

>>> result
(-2.6825663173365015, 0.077103947319183241, 0, 7, {'5%': -3.4775828571428571, '1%': -4.9386902332361515, '10%': -2.8438679591836733}, 15.971188911270618)

I'm thinking there is some built-in determining factor for the number of lags based off my input for the maxlag parameter, but I wasn't able to find anything online.

Could someone please explain this, and possibly how to make the ADF test run with the number of lags I want to input.

Answer 1

maxlag is the largest lag that is used for the autolag search.

By default the number of lags that are chosen is determined by the smallest AIC for all lags up to maxlag. If you want to have a fixed number of lags, then you need to use autolag=None as argument.

see the docstring http://www.statsmodels.org/dev/generated/statsmodels.tsa.stattools.adfuller.html