I have a model trained on my data. I used joblib to get the model and shared with other teams to evaluate the performance of the model on their data. One of the team came back and said that the models are not very well calibrated and so I thought of suggesting them to apply isotonic regression. The team only have 1 dataset which they do not want to split. Shall I use prefit (as below) and will it lead to overfitting or any other problem? Is there any other way to apply isotonic regression? I also don't want to apply isotonic regression on my model and then send it to them because the model is working fine for other teams.
calibrated_rf_isotonic_prefit = CalibratedClassifierCV(base_estimator=pretrained_model, method='isotonic', cv='prefit')
calibrated_rf_isotonic_prefit.fit(X, y)
