Pyramid is a no-nonsense statistical Python library with a solitary objective: bring R's auto.arima functionality to Python. Pyramid operates by wrapping statsmodels.tsa.ARIMA and statsmodels.tsa.statespace.SARIMAX into one estimator class and creating a more user-friendly estimator interface for programmers familiar with scikit-learn.

Pyramid is on pypi under the package name pyramid-arima and can be downloaded via pip:

$ pip install pyramid-arima

To ensure the package was built correctly, import the following module in python:

from pyramid.arima import auto_arima

For an easy, reproducible quick-start example, see examples/.

• How do I make pyramid run as quickly as R?

  • R's code is heavily C-based. Pyramid runs on statsmodels, which is Python based. There will be some differences in performance speed-wise, but much of it can be eliminated by using stepwise=True. See this discussion for more thoughts...

• Refreshing ARIMA models

  • Periodically, your ARIMA will need to be refreshed given new observations. See this discussion and this discussion on either re-using auto_arima-estimated order terms or re-fitting altogether.