I am trying to create a prediction using the "I have historical data" option as a training exercise for a workshop I'm leading, and I'm running into two issues.

According to the below documentation, the historical data set must include a value for all model context properties as well as pyOutcome. However, those properties aren't available to the Customer context, which means it doesn't seem possible to use the "I have historical data" option when a prediction is built in this context. Attempting to do so throws an error that causes all historical records to be ignored.

When a prediction is built in the Decision Request context, pyOutcome isn't available. In this case, the outcome of the historical data evaluation produces no cases for any outcome in the data set, because they aren't recognized.

The second issue when that when predictions are built in the Decision Request context, but the predictors that I would like to use are Customer data, those predictors aren't available when the prediction is built and trained on the historical data. (The Decision Request context has been required when I've built predictions in the past. At that time, I did not need to train the associated model, but if I did, I would have needed to use some other method.)

Is there a way to work around these issues? Am I using the "I have historical data" option correctly?

Thank you,

Cheri

Documentation: https://community.pega.com/knowledgebase/articles/decision-management/86/creating-predictions-customer-engagement