Skip to content

JoffreyDumontLeBrazidec/ibisa

Repository files navigation

DOI

IBISA

The goal of IBISA (Inverse Bayesian Inference for Source Assessment) is to provide an algorithm to sample radionuclide release sources.

Installation

Use the command

scons

in the main directory to compile and create an executable "mcmc"

Usage

./mcmc myconfiguration/config-conc.cfg

 Description

This repository is organised as follows:

Each class is located in a file with the same name, in a folder with the same name. Each folder is described below:

  • configuration: contains a typical configuration file (in particular, paths must be changed).

  • controller: contains a list of std::exception() to check the validity of each configuration parameter and facilitate bug recovery.

  • state: contains MarkovState class used to treat information about a Markov state (stage of the Markov chain)

  • cost: contains Cost class used to calculate the cost relative to a Markov state.

  • managers: contains the managers store, alpha, and cycle used to organise the storage of the MCMC results, the acceptance ratios of the MCMC algorithm, and the update cycle of the Markov chain, respectively.

  • sortingobs: contains the sortingObs class used to organise the sorting of observations into relevant or irrelevant.

  • observationOperator: contains the observationOperator class used to load, hold, and return information about the observation operator matrix.

  • redoperator: contains the ReducDim_Obs_Operator class called by observationOperator and used instead when transdimensional walks are allowed.

  • storage: contains the Storage class used to store the results of the reversible jump MCMC algorithm.

  • main: contains the paralleltempering.cpp file used to call the MCMC reversible jump algorithm and describes its cycle.

 Authors and acknowledgment

This code goes along the article:
"Bayesian transdimensional inverse reconstruction of the 137Cs Fukushima-Daiichi release"
Joffrey Dumont Le Brazidec, Olivier Saunier, Marc Bocquet, and Yelva Roustan

It was developed as part of a PhD at
IRSN (Institut de radioprotection et de sûreté nucléaire), Fontenay-aux-Roses, France
CEREA (Centre d'Enseignement et de Recherche en Environnement Atmosphérique) at the Ecole des Ponts and EDF R&D, Ile de France, France

 Support

Contact: joffrey.dumont@enpc.fr

About

Inverse Bayesian Inference for Source Assesment

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages