0

I am estimating ARIMAX and SARIMAX models (time series analysis), and I have some questions about their Python implementation within the statsmodels package.

  1. The ARIMA class in statsmodels is supposed to be able to also deal with SARIMAX models ("This model is the basic interface for ARIMA-type models, including those with exogenous regressors and those with seasonal components. The most general form of the model is SARIMAX(p, d, q)x(P, D, Q, s)."). However, statsmodels also offers a specialized SARIMAX class. What are the differences between the two? This FAQ page seems to offer some explanations, but I'm not sure I exactly understand the core of the issue.

  2. If I run a SARIMAX model in statsmodels (with either of the two classes above), and the order of integration d is nonzero, will the exogenous regressors be differenced as well?

  3. Again, if I run a SARIMAX model in statsmodels and the order of integration d is nonzero, is the regression of the target variable on the exogenous variables run before or after differencing? In general, is it run at the beginning of the estimation procedure?

Thanks in advance,

Fede

Federico Bindi
  • 1
  • 1

0 Answers0