Abinand and I were discussing a version of simulateNewDataFun that uses the model instead of a provided function, particularly for scenarios where a user has a complicated model and simulating data may not be straightforward. Could the default perhaps be to simulate data with the model + posterior sample of the top-level nodes in the model? Or the model + posterior sample of nodes in paramNames?
Abinand and I were discussing a version of
simulateNewDataFunthat uses the model instead of a provided function, particularly for scenarios where a user has a complicated model and simulating data may not be straightforward. Could the default perhaps be to simulate data with the model + posterior sample of the top-level nodes in the model? Or the model + posterior sample of nodes in paramNames?