DRlib: a C++ library for marked Gibbs point processes simulation and inference
Résumé
DRlib is a new C++ library for modeling, simulation and inference of marked point processes. The aim is to complete existing tools such as the spatstat library in R [1] with reliable and efficient C++ code allowing intensive Bayesian MCMC based inference for marked Gibbs point processes. This proposal has its roots in the MPPLIB library developed mainly by [2] where exact simulation algorithms for marked point processes were translated into C++ in order to perform massive simulation studies. The current version of the library is freely available at the following address https://gitlab.univ-lorraine.fr/labos/iecl/drlib.
Domaines
Probabilités [math.PR]Origine | Fichiers produits par l'(les) auteur(s) |
---|