Initialization of the Gibbs Algorithm according to the fixed stratigraphy
Source:R/Gibbs.R
initialize_SC.RdIn this function we try different way of initializing our Gibbs Sampler. Several limitation can be encoutered :
The constraints need to be respected
The ages need to belongs to a certain period of study T1, T2 that will be fixed arbitrarily
Initialisation are not only constrained by the Matrix of constraint but also to avoid numerical zeros the interval of initialisation need to be reasonable according to the density of Gibbs Sampler.
The most efficient approach would be to considered the approximation Ahat = Dhat / ddot and then make it respect the constraints