Fill gaps from assistant conversations: CV cutoffs, SARIMAX, plots

[mintlify]

Summary

Answers recurring assistant questions directly in the docs: how cross-validation windows and cutoffs are determined, where SARIMAX functionality lives in StatsForecast, and how to customize plot_series output.

Changes

• Cross-validation tutorial: note that h/step_size are measured in timesteps of the data's frequency, plus a new "How cutoffs are determined" section explaining that cutoffs are anchored to the end of df and the trim-and-rerun workaround for exact cutoff dates.
• ARIMA model page: note that there is no separate SARIMAX class; ARIMA/AutoARIMA with season_length and exogenous variables via X_df cover it, linking to the exogenous variables guide.
• utilsforecast plotting page: new section showing how to set titles and axis labels on the returned matplotlib/plotly figure, since plot_series has no title/label parameters.

Context

Assistant conversation review (past week) showed repeated questions on step_size units and setting explicit cutoff dates, searches for a SARIMAX model, and questions about customizing plot titles/axes.

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated
nixtlaverse_nixtla_io	🟢 Ready	View Preview	Sep 7, 2026, 4:15 PM

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}