To start the uncertainty quantification with the Uptimai software package, a project has to be started first. This first step can be done easily using the Uptimai Launcher program which can be found with the main folder of the Uptimai software installation. When the working project is set with its path, other tools of the software package may be used. They are accessible via buttons of the Uptimai Launcher window.
The initial state of the application's window can be seen in Figure 1. At the very
top, there is a Menu bar with an ordinary arrangement of items. Under
the File menu, a completely new project can be started as the New
project by selecting an existing folder, which name becomes the name of
the project. Creation of a new project is finished when the program
*project_name*.uptim file into the project folder. Load
project function allows to continue in an already existing project by
finding and selecting its
*.uptim file. About item of the Menu bar
contains features accessing an off-line version of this document (Help) and company
information and contact (Company).
- Terminal launcher : Select your preferable terminal launcher which will run the Core Solver. Launchers available for your machine are found automatically and listed in the combobox.
- Input Preprocess : Creates input distributions for the uncertainty quantification process. It also allows to load already built distributions, plot them and export graphic outputs.
- Core Solver Setup : Sets all parameters of the Core Solver together with directions where to store final data of the uncertainty quantification.
- Run Core Solver
: Performs the actual uncertainty quantification
and stores results into
*.jsonfiles. Requires all files created in Input Preprocess and Core Solver Setup modules and needs to be coupled with the code which evaluates the output of the expensive function.
- Result Postprocess : The tool to analyze the result of the uncertainty quantification thoroughly. Computes the sensitivity analysis, creates plots of increment functions, their histograms and influencer plots as well as data convergence history. All outputs can be stored for further use.