What’s new in Pyramid

As new releases of Pyramid are pushed out, the following list (introduced in v0.8.1) will document the latest features.


  • Explicitly catch case in auto_arima where a value of m that is too large may over-estimate D, causing the time series to be differenced down to an empty array. This is now handled by raising a separate error for this case that better explains what happened.
  • Re-pickling an ARIMA will no longer remove the location on disk of the cached statsmodels ARIMA models. Older versions encountered an issue where an older version of the model would be reinstated and immediately fail due to an OSError since the cached state no longer existed. This means that a user must be very intentional about clearing out the pyramid cache over time.
  • Added pyramid cache check on initial import to warn user if the cache size has grown too large.
  • If d or D are explicitly defined for auto_arima (rather than None), do not raise an error if they exceed max_d or max_D, respectively.
  • Added Circle CI for validating PyPy builds (rather than CPython)
  • Deploy python wheel for version 3.6 on Linux and Windows
  • Include warning for upcoming package name change (pmdarima).


  • ARIMA instance attributes

    • The pkg_version_ attribute (assigned on model fit) is new as of version 0.8.0. On unpickling, if the current Pyramid version does not match the version under which it was serialized, a UserWarning will be raised.
  • Addition of the _config.py file at the top-level of the package

  • Fix bug (Issue #30) in ARIMA where using CV with differencing and no seasonality caused a dim mismatch in the model’s exog array and its endog array

  • New dataset: Woolyrnq (from R’s forecast package).

  • Visualization utilities available at the top level of the package:

    • plot_acf
    • plot_pacf
    • autocorr_plot
  • Updated documentation with significantly more examples and API references.


  • ARIMA out_of_sample_size behavior
    • In prior versions, the out_of_sample_size (OOSS) parameter misbehaved in the sense that it ended up fitting the model on the entire sample, and scoring the number specified. This behavior changed in v0.7.0. Going forward, when OOSS is not None, ARIMA models will be fit on \(n - OOSS\) samples, scored on the last OOSS samples, and the held-out samples are then added to the model.
  • ARIMA add_new_samples method
    • This method adds new samples to the model, effectively refreshing the point from which it creates new forecasts without impacting the model parameters.
  • ARIMA confidence intervals on predict
    • When return_conf_int is true, the confidence intervals will now be returned with the forecasts.


  • CHTest of seasonality
    • No longer compute the \(U\) or \(V\) matrix in the SVD computation in the Canova-Hansen test. This makes the test much faster.