TabularPredictor.predict_oof

TabularPredictor.predict_oof(model: str | None = None, *, transformed=False, train_data=None, internal_oof=False, decision_threshold=None, can_infer=None) → Series[source]

Note: This is advanced functionality not intended for normal usage.

Returns the out-of-fold (OOF) predictions for every row in the training data.

For more information, refer to predict_proba_oof() documentation.

Parameters:

  • model (str (optional)) – Refer to predict_proba_oof() documentation.

  • transformed (bool, default = False) – Refer to predict_proba_oof() documentation.

  • train_data (pd.DataFrame, default = None) – Refer to predict_proba_oof() documentation.

  • internal_oof (bool, default = False) – Refer to predict_proba_oof() documentation.

  • decision_threshold (float, default = None) – Refer to predict_multi documentation.

  • can_infer (bool, default = None) – Refer to predict_proba_oof() documentation.

Return type:

pd.Series object of the out-of-fold training predictions of the model.