Getting and installing UQLab
This page gathers information to install and configure UQLab, the Framework for Uncertainty Quantification.
Obtain the source code
As of version 2.0.0, UQLab has gone completely open source, and no licensing is required anymore.
It is now sufficient to register on this website to have access to the full UQLab source code.
You can easily register/login by clicking on the login link at the top-right of the page, or directly at the following link: Login/Register ➜
After you login, you will find the download link for the latest version of UQLab directly in your personal page ➜
Note for existing UQLab users that were already registered with an academic/commercial license:
Your previous registration has been automatically imported into the new system. Please use the "forgot password" link to setup a login password on the website
Unzip the downloaded zip archive in a folder, say my_UQLab_folder
Open MATLAB and select my_UQLab_folder/core as your working directory
The minimal system requirements to run UQLab are:
MATLAB® R2017a and later
Statistics and Machine learning toolbox ➜ (formerly "Statistics toolbox")
To access all the functionalities, the following toolboxes are required:
Optimization toolbox ➜ - Kriging, RBDO, SVR, SVC, Copula Inference (BFGS)
Global Optimization toolbox ➜ - Kriging, RBDO, SVR, SVC (BFGS) (global optimizers)
MATLAB® is a registered trademark of The MathWorks, Inc.
Description of the installation subfolders
After installation, the UQLab code is organized in several subfolders:
/core: contains the core routines including the memory/session handling and utilities.
/Doc: contains the UQLab user manuals in PDF format. Each module (namely Input, Model, Inference, Inversion, Kriging, Polynomial Chaos, PC-Kriging, LRA, SSE, SVC, SVR, Sensitivity, Reliability, Random fields, RBDO, UQLink, and UQLib) has its own user manual.
/Examples: contains example files gathered in subfolders according to the type of analysis addressed. By using MATLAB’s built-in publish function you can transform these example files into formatted,
self-contained HTML files for better readability.
/lib: contains shared mathematical routines required by the various modules (UQLib).
/modules: contains the open source scientific code of the various modules.
/Profiles: contains the templates for the HPC dispatcher module remote machine profile file
You can test your UQLab installation by typing uqlab -selftest once UQLab has been installed using the above procedure. A full suite of self-test routines is run (this may take up to several minutes on a standard computer). If everything is OK, a message will confirm that the tests were successful.
Still cannot install properly?
Please don't hesitate to contact us ➜
We will be glad to assist you in the installation or troubleshooting process.