If the function runGompitzPrediction is applied to a subset of data that does not cover all cohorts included in the calibration object, stratum names are allocated falsely in returned object.
There are several ways of fixing this:
- delete relevant stratum from calibration object
- Include fake data for all cohorts in network data and delete after prediction
- just correct allocation of stratum names, maybe via manipulating the stratum matrix
If the function
runGompitzPredictionis applied to a subset of data that does not cover all cohorts included in the calibration object, stratum names are allocated falsely in returned object.There are several ways of fixing this: