Bayesian Subset Simulation (BSS)

This repository provides an implementation of the "Bayesian Subset Simulation" (BSS) algorithm described in the following preprint:

Julien Bect, Ling Li & Emmanuel Vazquez
Bayesian Subset Simulation
SIAM/ASA Journal on Uncertainty Quantification 5(1):762–786
DOI:10.1137/16M1078276 / arXiv:1601.02557 / HAL:01253706

It is currently in a very preliminary form. After some polishing, it will (sooner or later) become part of the STK toolbox.

Requirements

You need an interpreter of the Matlab/Octave langage, that is, either Mathworks's Matlab or GNU Octave.

The result presented in the article were obtained in 2017 using Matlab R2014b and STK 2.4.0. Please report any problem with other versions of Matlab or with Octave.

Since BSS is meant to become part of the STK toolbox, it should ideally run with Matlab >= R2009b and Octave >= 4.0.0 (those are the current requirements indicated in STK's README). Right now, it probably doesn't...

Installation

1. Clone the BSS repository wherever you like:

   git clone [email protected]:stk-kriging/contrib-bss.git bss
   

   (or download and unpack a zip snapshot of the repo).

2. Download and unpack STK 2.4.0 (or any later version) in bss/stk. If you don't know which tarball to pick, stk-*-allpurpose.tar.gz is the one you're looking for. Alternatively, Octave users can prefer to install stk as a regular Octave package (pkg install -forge stk), in which case bss_init.m must be slightly modified.

3. Start Matlab or Octave and then type

   bss_init
   

   to initialize BSS (and STK). That's all, you're ready to use BSS.

Test

You can test that BSS works as expected by running the test script:

script_test_bss

The results are summarized at the end in a table that should look like this:

case #001:  DIM=6  pf_estim=1.05e-05  pf_ref=9.09e-06  ratio=1.15e+00  
case #002:  DIM=6  pf_estim=1.77e-08  pf_ref=1.56e-08  ratio=1.13e+00  
case #003:  DIM=2  pf_estim=4.15e-06  pf_ref=3.94e-06  ratio=1.05e+00  
case #004:  DIM=2  pf_estim=7.19e-09  pf_ref=5.49e-09  ratio=1.31e+00  
case #005:  DIM=1  pf_estim=8.26e-06  pf_ref=1.00e-05  ratio=8.26e-01  
case #006:  DIM=1  pf_estim=1.07e-05  pf_ref=1.00e-05  ratio=1.07e+00  
case #101:  DIM=2  pf_estim=1.37e-04  pf_ref=2.03e-04  ratio=6.73e-01 

The ratios reported in the last column should be reasonably close to one.

Repository

The current home of BBS is on github: https://github.com/stk-kriging/contrib-bss

A mirror in also maintained at Sourceforge: https://sourceforge.net/p/kriging/contrib-bss (this is the URL that was given in the published paper).

Please use the github project if you want to fork BSS, report issues, or whatever.

Licence

Copyright (C) 2016, 2017 CentraleSupelec

This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version.

This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA