The new Persalys version is available, do not hesitate to ask it.
Here is the list of changes:
- Improvement of the coupling model : There are new input and output tests to help the user to check the validity at each step of the coupling procedure. The user can choose now the file format, and a summary tab is provided where all inputs and outputs are listed. Within the coupling model, it is possible now to add a Python model in order to perform mathematical pre/post operations.
- New design of experiments : An optimized Latin Hypercube sampling is proposed to the user.
- Data analysis : The user can now load sparse data as a cleaning wizard is integrated into the data model to allow the user to choose how to clean the sample (delete or replace the missing values).
- aggregated Sobol indices are also provided when several outputs are defined in the physical model.
- Python editor : The usability of the editor has been improved with syntax highlighting and zoom options in the text.
- Parallel evaluations : The parallelization is now process based which allows the user to exploit all available CPUs, which can be important especially for a Python model.
- And of course we fixed some bugs.