Categories: Optimisation, Surrogate Modeling
State: ❌ must be refactored to work with the new version of the code (see Issue #147)
SMTrain is a module to train a surrogate model, with data calculated with CEASIOMpy.
SMTrain takes as input a CPACS file, it must contained a path to a CSV file with the training dataset in it, or an aeromap can be chosen instead. However the aeromap must have the outputs (cl, cd, ...).
SMTrain module generates a surrogate model by using a dataset which contains a list of inputs with their corresponding outputs. Therefore any parameters can be given to create a surrogate.
SMTrain outputs a CPACS files with the new values set by the user. It will be ready to be used by the next module in the workflow.
SMTrain is a native CEASIOMpy module, hence it is available and installed by default. It is base on the SMT: Surrogate Modeling Toolbox
For now the specific training options for a model are not implemented in the GUI and must be modified within the code.