User Guide

The following guides cover how to get started with a pmdarima distribution. The easiest solution is simply installing from PyPi, but if you’d like to contribute you’ll need to be able to build from source, as laid out in the Setup section.

• 1. About the project
  • 1.1. The name…
  • 1.2. How it works
  • 1.3. Feedback
• 2. Setup
  • 2.1. Install from PyPi
  • 2.2. Build from source
• 3. Quickstart
  • 3.1. Auto-ARIMA example
• 4. Serializing your ARIMA models
• 5. Refreshing your ARIMA models
  • 5.1. Updating your model with new observations
• 6. Tips to using auto_arima
  • 6.1. Understand p, d, and q
  • 6.2. Understand P, D, Q and m
  • 6.3. Parallel vs. stepwise
  • 6.4. Pipelining
• 7. When no viable models can be found
• 8. Encountering issues in seasonal differencing
• 9. Toy time-series datasets
  • 9.1. Endogenous Datasets
  • 9.2. Exogenous Datasets
• 10. Use cases
  • 10.1. Stock Market Prediction
  • 10.2. An end-to-end time series analysis
• 11. Contributing to pmdarima
  • 11.1. How to contribute
  • 11.2. Pull Request Checklist
  • 11.3. Filing a bug
• 12. Contributors
• 13. Citing