Incorrect behavior on updating versioned Adaptive model
On updating an Adaptive model in a branch after versioning the model, the changes made display on the UI. However, the changes are not saved to the database.
Steps to Reproduce
- Open the Pega Prediction Studio.
- Open the Adaptive Model rule that must be updated.
- Navigate to Actions > Save As. Note the Predictor Types of the first page of predictors.
- Select a Development branch.
- Click 'Create and open'.
A defect in Pegasystems’ code or rules: Adaptive Model rule UI in both Prediction Studio and Dev Studio.
When a non-default Predictor type is selected, the change made displays on the UI, but is not saved to the database. An error does not display during the Save. Hence, incorrect data is stored. That is, when data has accrued, the rule cannot be Saved or Saved As because Validation starts, believing that the Predictor type of a Predictor with data is changing.
Perform the following local-change:
- Create a new Adaptive Model rule (or Private Edit an existing rule).
- Add a field Predictor based on a property of the class. On selection, the property type and Predictor type of the row are filled with the default values.
- Change the Predictor type to another available value.
- View the XML of the rule through the Actions menu. Confirm that the new Predictor is in both the pyStaticPredictors and pyPredictors list elements and that the value for the pyPredictorType element is the same as on the UI.