Persalys offers the possibility to realize data analysis by creating data models. You can import a text file containing a values matrix of the measured parameters that may be considered as inputs or outputs. It can be a spare sample, a cleaning wizard is provided to allow you to decide if you want to delete these lines or to replace it with a specific value.
It is possible to perform several analyses on these data:
- perform a descriptive and visual statistical analysis of the data
- infer 1-D probabilistic distributions on each of the parameters
- infer a dependency model to consider the correlation between parameters
- build metamodel allowing to define a mathematical model that will link the inputs with the outputs. The metamodel can then be converted into a physical model to use the various probabilistic analyses
For more information on the inference of distributions from data, refer to the following video (soon in English):
For more information on building a metamodel from data, refer to the following video (soon in English):