regression_model.RdEta is a matrix of individual-level parmaters for a regression-spline with equivalent knots over all individuals
regression_model(observation_model_parameters, MegaLMM_state = list())a list including:
1) observations, a data.frame with observation-level data including columns ID and Y
3) individual_model the model that should be applied to the data for each ID.
The current MegaLMM_state.
For initialization, can be a list with data_matrices = list(data=data)
Where data contains ID for ordering.
This function should pre-calculate design matrices for each individual (assuming they are all
unique). During sampling, should sample regression coefficients Eta given the factor model state.
Set up to allow for non iid errors, but currently not implemented.