Changes That Invalidate Pipelines

Some changes that you make to your project can make your pipelines out of date. When this happens, the pipelines must be run again to pick up the changes. All pipelines must be run again to correctly reflect any of the following changes:

• changing the forecasta numerical prediction of a future value for a specified time period for each unique combination of BY variable values settings in Project Settings.
• changing or refreshing the data source for a time seriesan aggregation of transactional data into specified time intervals and sorted according to unique combinations of the default attributes (BY variables) or external forecast project, or imported attributes. Any change to the data source must follow the same schema as the original.
• adding attributes to the project using the Attributes view of the Data tab.
• adding or removing any independent variables or changing their settings.
• any changes to eventsan incident that disrupts the normal flow of any process that generates the time series. Examples of events are holidays, retail promotions, and natural disasters. in the project.
• changing the modeling nodes in a pipeline. This includes adding or removing nodes, or changing certain options for a node. Only the affected pipeline needs to be run again.

In some circumstances, you might need to invalidate a pipeline so that you can run it again and pick up new changes. In this case, you can change one of these settings and then reset it to the previous setting. The status of the nodes indicates Out-of-date. You can run the pipeline again to pick up the latest changes.

After you run a pipeline and open the Overrides tab, variable assignments and attribute properties are committed and cannot be changed in the project. For more information, see Locking the Data Definition.

If you need to make any changes to the committed settings, for example, setting a different time interval for the time variable, you can start a new project using the same data source.