Error "The data is too short for the required differences." Occurs When Forecasting Monthly Metrics Using Time Series Forecasting (ARIMA/SARIMA)

When performing monthly metrics forecasting using Time Series Forasting (ARIMA/SARIMA), you may encounter the error message “The data is too short for the required differences.” and receive no results.

This error occurs when the aggregated data spans less than three months. (If using the “Repeat By” option, it means categories selected for “Repeat By” contains data that meets this condition.)

To resolve this issue, please extend the time period of your data.

Additionally, if you are using the “Repeat By” option, you can avoid the error by removing categories with a short data period from the dataset in advance.