Iterative File – Sample Data

Table of Contents

During the uncertainty quantification process, the Core Solver stores the progress of each iteration into the *.json file. It contains combinations of input variables and function value of each evaluated sample as well as names of input variables, boundaries of the domain, etc. In the Iterative File method one can verify function value of each sample used for the UQ process and, if necessary, make corrections. It is also possible to add new samples into the dataset, for all samples can be added another output too.

How to use it: #

Fig. 1: Iterative File – File loaded

The *.json file to be reviewed/edited needs to be loaded via File ->Load, then can be the method activated. Data from the file are presented in a form of a spreadsheet. To change values in cells, proceed the same way as in any type of spreadsheet editor. Remember, that only numbers are allowed in cells of the iterative file. These also need to be inside the computational domain. Thus, they can’t be lower or higher than boundaries defined in the InputDomain.txt file.
Many options of the table can be accessed by right-clicking on it. Right-click on a row number opens the context menu allowing to Delete or Add row(s). Right-click on a row number opens the context with Column -> Add new column. Due to practical reasons, it is not recommended to use other options featured with the Pandas library used to create the spreadsheet. When adding rows (new samples) or columns (additional outputs), do not forget to fill all cells with data. Formatting of the iterative *.json file does not support blank spaces.

Buttons: #

  • Check: Helps to identify wrong or empty inputs in the table. When incorrect input is found, the user is informed about the source of an error. Saving is disabled until all inputs are correct.
  • Save: Saves changes made into the iterative *.json file.