train_disp_vector

lstchain.reco.dl1_to_dl2.train_disp_vector(train, custom_config=None, predict_features=None)

Train a model (Random Forest Regressor) for the regression of the disp_norm vector coordinates dx,dy. Therefore, the model must be able to be applied on a vector of features. TODO: introduce the possibility to use another model

Parameters:

train: `pandas.DataFrame`

custom_config: dictionnary

Modified configuration to update the standard one. Default=None

predict_features: list

list of predict features names. Default=[‘disp_dx’, ‘disp_dy’]

Returns:

The trained model