What is MUQ?

In a nutshell, MUQ is a collection of tools for constructing models and a collection of uncertainty quantification (UQ)–focused algorithms for working on those models. Our goal is to provide an easy and clean way to set up and efficiently solve UQ problems. On the modelling side, we have a suite of tools for:

  • Combining many simple model components into a single sophisticated model.
  • Propagating derivative information through sophisticated models.
  • Integrating ordinary differential equations and differential algebraic equations (via Sundials)

Furthermore, on the algorithmic side, we have tools for

  • Performing Markov chain Monte Carlo (MCMC) sampling
  • Constructing polynomial chaos expansions (PCE)
  • Computing Karhunen-Loeve expansions
  • Building optimal transport maps
  • Solving nonlinear constrained optimization problems (both internally and through NLOPT)
  • Regression (including Gaussian process regression)

Installation

Using Docker

docker pull mparno/muq2
docker run -it --rm mparno/muq2 bash

From source

Currently, the best way to get started is by checking out our git repository on bitbucket. Go to the folder you where you want to keep the MUQ source code, and use:

git clone https://bitbucket.org/mituq/muq2.git

Compiling the source

Now that you have the source, you need to compile and install MUQ. In a very basic installation of MUQ, all you need to do is specify an installation prefix. To keep all of the installed MUQ files together, we suggest using something other than /usr/local for the prefix. A typically choice may be ~/Installations/MUQ_INSTALL. For this basic installation of MUQ, cd into muq2/build/ and type:

cmake -DCMAKE_INSTALL_PREFIX=/your/install/directory ../

During this command, cmake will generate a make file, which can now be run as usual:

make -j4 install

The -j4 is an option specifying that make can use 4 threads for parallel compilation.

How can I learn how to use MUQ?

  • Look at the examples listed on our examples page.
  • Try out some of our examples by launching a temporary interactive MUQ session. No installation necessary!
  • Check out the doxygen documentation here.
  • Join our slack workspace and ask a question.
Slack LOGO MUQ is on Slack!
  • Join our slack workspace to connect with other users, get help, and discuss new features.
Test Status
Acknowledgments
NSF Logo This material is based upon work supported by the National Science Foundation under Grant No. 1550487.

DOE Logo This work was supported by the DOE Office of Science through the QUEST and FastMath SciDAC Institutes.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.