The QUESO Library, User's Manual

TL;DR

QUESO is a flexible, portable C++ library designed for uncertainty quantification in mathematical modeling, supporting inverse and forward problems, model validation, and prediction with uncertainty analysis.

Contribution

This manual introduces QUESO's object-oriented design, its capabilities for uncertainty quantification, and provides usage procedures and example studies.

Findings

Supports statistical inverse and forward problems

Enables model validation under uncertainty

Provides tools for uncertainty-based prediction

Abstract

QUESO stands for Quantification of Uncertainty for Estimation, Simulation and Optimization and consists of algorithms and C++ classes intended for research in uncertainty quantification, including the solution of statistical inverse problem and statistical forward problems, the validation of mathematical models under uncertainty, and the prediction of quantities of interest from such models along with quantification of their uncertainties. QUESO is designed for flexibility, portability, ease of use and ease of extension. Its software design follows an object oriented approach and its code is written on C++ and over MPI. It can run over uniprocessor and multiprocessor environments. QUESO contains two forms of documentation: a user's manual available in PDF format and a lower level code documentation available in web based/HTML format. The present document is a user's manual which provides an overview of the capabilities of QUESO, procedures for software execution, and includes example studies.