Plotting prior and posterior predictive sampling

[LegrandNico]

Sampling the HGF could mean different things. For prior predictive sampling, we could try to implement:

• Predictive sampling from the generative model using the probabilistic graph (starting from the leaves up to the roots). This would require writing sampling update steps and a sampling update sequence. The function only requires a number of time step as input argument.
• Prior and posterior predictive sampling from a distribution over a parameter given the input data. In that case, we just run the model forward using different parameters.

Both samplings should interface with a dedicated plotting function to visualize trajectories.