Statsmodels Time Series

First published: 2009

Brief summary

Python docs for ARIMA, state-space and decomposition.

Article

Statsmodels Time Series is covered by the linked source. Python docs for ARIMA, state-space and decomposition.

The material from Statsmodels presents a forecasting method, software implementation, benchmark or evaluation procedure.

Sound forecasting preserves time order, compares models with simple baselines and evaluates performance at the intended horizon. Point accuracy and uncertainty may both matter.

Performance can deteriorate after structural change. Model selection and preprocessing must use training data only, with final results measured on genuinely unseen periods.

Source details and credits

• Source / publisher: Statsmodels
• Source type: Software docs
• URL type: WWW
• Credits: Statsmodels
• URL: https://www.statsmodels.org/stable/tsa.html