| train {GEARS} | R Documentation |
train(gears);the training epoch numbers and the learning rate of the model training is determined by the GEARS_opts configuration object that is given at the model creation time, and the loss value of each training epoch is stored in the LossCurve property of the trained model object.
an error will be thrown if the training sample set is empty, i.e. there is no gene of the prior network could be mapped into the gene expression matrix.
the input GEARS model object that has been trained, which can be used for run the in silico gene perturbation experiment via the knockouts, overexpress and knockdown api of the bnlearn package module.