In case multiple thresholds give the same error the largest one is chosen (i.e. the one keeping the fewest features).
predict_pamr(object, x, threshold, thres_fun, ...)Fitted classifier.
Dataset of observations to be classified.
Threshold to use for classification. This argument is only needed if you want to override the value set during model fitting.
Threshold selection function. Only needed if you want to override the function set during model fitting.
Sent to pamr.predict.
A list with elements:
prediction: Factor of predicted class memberships.
probability: Data frame of predicted class probabilities.