DREAM_LOA: Limits of Acceptability with DREAM algorithm: MATLAB and Python code

Description

DiffeRential Evolution Adaptive Metropolis (DREAM) algorithm with Limits of Acceptability (LOA) sampling. The MATLAB and Python toolboxes delineate the behavioural solution space of set-theoretic likelihood functions used within the Generalized Likelihood Uncertainty Estimation (GLUE) LOA framework (Beven 1992, 2001, 2006, 2014). The algorithm builds on the DREAM$_{(ABC)}$ algorithm of Sadegh and Vrugt (2014) and enhances significantly the accuracy and CPU-efficiency of Bayesian inference with GLUE. In particular it is shown how lack of adequate sampling in the model space might lead to unjustified model rejection.

Getting Started

Installing: MATLAB

• Download and unzip the zip file 'MATLAB_code_DREAM_LOA_V1.0.zip' in a directory 'DREAM_LOA'
• Add the toolbox to your MATLAB search path by running the script 'install_DREAM_LOA.m' available in the root directory
• You are ready to run the examples.

Executing program

• After intalling, you can simply direct to each example folder and execute the local 'example_X.m' file.
• Please make sure you read carefully the instructions (i.e., green comments) in 'install_DREAM_LOA.m'

Installing: Python

• Download and unzip the zip file 'Python_code_DREAM_LOA_V1.0.zip' to a directory called 'DREAM_LOA'.

Executing program

• Go to Command Prompt and directory of example_X in the root of 'DREAM_LOA'
• Now you can execute this example by typing 'python example_X.py'
• Instructions can be found in the file 'DREAM_LOA.py'

Authors

• Vrugt, Jasper A. (jasper@uci.edu)

Version History

• 1.0
  • Initial Release
  • Python implementation

Acknowledgments