Draft: Joint model with multiple longitudinal outcomes
ON PAUSE, WAITING FOR MR !106 (closed) TO BE MERGED!
I have created a new merge request because the other one deviates too far from its original branch.
The idea of the joint model with multiple longitudinal outcomes could be separated into two distinct MR to make it easier to review. A first MR, this one, whose goal is to create a joint model that inherits from a multivariate model but that still has the survival observation model from the univariate model. This would be an under-parametrised model. Then once this is done, the idea will be in another MR to add the proper observational model to enable a better parametrization of the structure.
Note that a conversation will be needed to decide if we should keep the joint_univariate AND the joint_multivariate.
What does the code in the MR do?
The MR will create a new model, that enables different observation models for the multivariate model and one for the survival model, but that enables the user to give whatever combination of both ('surv_obs', 'long_obs'), ('long_obs', 'surv_obs').
Where should the reviewer start?
WIP
How can the code be tested?
WIP