td-portman

What problem does it solve?

This Skill automates the process of performing Ljung-Box portmanteau tests on time series model residuals, helping you diagnose model adequacy and identify autocorrelation.

Core Features & Use Cases

• Model Diagnostics: Apply Ljung-Box and Box-Pierce tests to assess if residuals behave like white noise.
• Autocorrelation Detection: Identify patterns of autocorrelation in model residuals, indicating potential model misspecification.
• Use Case: After fitting a time series model (e.g., ARIMA), use this Skill to test its residuals. If autocorrelation is detected, it signals that the model may need refinement, such as adding more lags or seasonal components.

Quick Start

Run the td-portman skill to diagnose the residuals from your fitted time series model.