Breaks

Breaks in a dataset are used to set a boundary between groups of data that are so far apart in time that you cannot sensibly interpolate between them. Breaks are used to prevent the product from finding any spurious time-dependent relationship between them when it trains a model or analyzes time-delayed data. Therefore, a Break is completely different from any other type of “bad” data. All transforms that operate on multiple rows of data treat Break as a boundary, not as a single bad data point. If you are using only user-specified models, the dataset values do not matter (except that the range of the data is used to set default variable bounds and constraints).
Provide Feedback
Have questions or feedback about this documentation? Please submit your feedback here.