Jumps Definition {evdbayes} | R Documentation |
These are function usefull to define reversible jumps for the algorithm. There is currently three functions: jacFun which evaluates the jacobian, movType1 which returns a special proposal of type 1 and movTyp2 which returns a special proposal of type 2.
jacFun(xi, pMass, cv) movTyp1(prow, pMass, cv) movTyp2(prow, propShape, cv)
xi |
The shape parameter at which the jacobian must be evaluated. |
pMass |
The point Mass i.e. the fixed value for the shape parameter. See reference guide. |
cv |
A numeric value for jumping rule. See reference guide. |
prow |
The current state of the Markov Chain. |
propShape |
The proposal shape candidate. |
These are special fucntions but can be overwritten to define new jumping rules.
See user's guide.
movTyp1
and movTyp2
returns a vector of lenght two. It
corresponds to location and scale proposals.
jacFun
return the value of the Jacobian at value xi
.
Mathieu Ribatet
Green, P. (1995) Reversible jump Markov chain Monte Carlo computation and Bayesian model determination. Biometrica 82:711–732.